Scorecard alignment

Person #2: Good morning everyone. Can you all hear me? I think Rashed is still joining. Person #3: Morning. Yes, we can hear you. I think it's just us and the Finance team. Person #21: Good morning. Yes, present. And Sameera is here with me. Person #20: Morning. Yes, good morning. Person #5: Morning. Can you hear me? I had some audio issues earlier. Person #3: Yes, we can hear you now. Person #6: Good morning, everyone. Yaani, I am here, but I think my mic is—can you hear me now? Person #2: Yes, we can hear you. Welcome. We're just waiting on the Chairman's office, but I think it's just the planning team and Finance today. So the quarterly KPI submission is due to ADEC by the fifteenth, and we need to align on the definitions before I send the final packet to the Undersecretary. I know we have some open items from the last review meeting. Person #3: Yes, we have the notes from last week. There were three items that Finance flagged. The savings calculation methodology, the local content threshold, and the entity compliance rate definition. And I think we also need to clarify the digital adoption metric from GovDigital. Person #21: Yes, the savings one is the main issue. Under the Standards, clause 4.2, we define savings as the difference between the baseline price and the awarded price, adjusted for volume. But the planning template we received from your team uses a different basis. It includes administrative savings, which we don't track. Person #3: With respect, the template was distributed in April and the definitions were in the annex. I recall we sent a clarification note in June. Did that not reach you? Person #21: It did. But the clarification note actually said the opposite of what the Standards say. So the teams are confused. They have two different definitions, and both claim to be authoritative. Person #5: Okay, so basically—sorry, can I just jump in? The digital adoption metric we submitted last quarter got bounced back by ADEC because the denominator was wrong. We used the civil servant headcount, but apparently they wanted the total active government service users. So we have a data issue there too. Person #2: Let me frame this. We have three, maybe four, definitional conflicts that need to be resolved today so the submission is consistent. I don't want the sectors submitting numbers that contradict each other. That was the problem in Q2. What I'd like to do is go through each disputed KPI, hear both sides, and agree on a single definition. And for the record, if the Standards conflict with the planning template, we need to flag that properly, not just pick one. Person #3: Understood. I have the list. Should I start with the procurement savings one? Person #2: Please. Person #3: So the KPI is "Government Procurement Savings". Target is seven hundred and fifty million dirhams for 2026. The definition in the planning template is: total savings realized across common categories, including price savings, demand management savings, and administrative efficiencies. The one in the procurement Standards is: savings from framework agreements in common categories, calculated as baseline price minus awarded price, volume-adjusted. So the main difference is the inclusion of administrative efficiencies, which the procurement side says they cannot evidence. Person #21: Correct. We cannot evidence administrative savings because they're not tracked in the Procurement Gate. We have the contract value, we have the baseline reference price, and we compare. That's it. The administrative part is, with all due respect, a concept from the consultants. Person #20: Yeah, the Deloitte team proposed it in the preliminary report, but when we actually tried to implement the tracking, the entities didn't record time savings or process efficiencies. So the data just isn't there. Person #5: But if we drop it, the target is basically impossible. Seven hundred and fifty million from price savings alone? The categories we have under the umbrella—ICT, facilities, vehicles—the baseline prices are already pretty optimized from last year's renegotiations. Person #21: The target was set before I joined this process. I did not inherit it. Person #3: The target came from the Undersecretary's office. And it was agreed with the Department of Finance in the autumn budget cycle. So we can't change the target. We can only change the definition. But if we change the definition to exclude administrative savings, we need to report the variance, and that will reflect on the Q3 scorecard. Person #6: If I may. When we had this issue in Riyadh, with the e-government program, we had a similar problem. The target was set by the ministry, but the definition was from a different era. So what we did—and it worked quite well, I have to say—is we kept the target, but we created a supplementary indicator that captured the administrative savings separately. So the main KPI was clean, and the supplementary one told the story. Person #2: That's one option. But it creates a two-tier reporting structure, which ADEC may or may not accept. They've been pushing for simpler scorecards, not more indicators. Person #5: They've been pushing for outcomes, not activity. So maybe that's the conversation we need to have. If we say, "Here is what we saved in dirhams, and here is the process improvement we enabled," that's outcomes. The administrative savings thing is more like—I mean, it's a result, but it's hard to verify. Person #3: For the record, ADEC's guidance says every KPI must have a single verifiable definition. So we can't have two definitions for the same indicator. We could have two separate indicators, but that increases the count from forty-seven to forty-eight, and the Chairman's office has been keen to keep it under fifty. Person #21: I understand. So what we need is a decision, not an analysis. If the Undersecretary wants the savings number to include administrative efficiencies, I will report it. But I will also add a footnote that the number is not verifiable under the current systems. Which, I think, weakens the whole submission. Person #2: Let's hold that thought. We're not going to resolve this in the next five minutes, and I'd rather we hear the other definitional issues first, then come back and make a single recommendation. What's the second one? Person #3: Local content. The KPI is "Percentage of procurement spending on local content suppliers". The definition dispute is around what counts as "local". The Standards say a local content supplier is registered with the Procurement Gate under the local content category and is majority-owned by UAE nationals. But the planning template says "suppliers with local manufacturing or assembly operations in the emirate". Person #21: Yes. That's a big difference. If we use the template definition, we bring in companies that have a warehouse in Khalifa Industrial Zone but are not majority-owned by UAE nationals. The local content mandate, which is a policy from the Executive Council, is specifically about national ownership. It's not about where the goods are assembled. Person #20: Also, the data is cleaner with the registration approach. We have a flag in the Gate. If we use the manufacturing definition, we would have to manually verify each supplier, and there are—I don't know—how many registered suppliers do we have now? Something like four thousand. Person #5: But wait. The Executive Council mandate is about national economic development. If a company with a factory in Abu Dhabi employing three hundred people—even if it's foreign-owned—if it's manufacturing locally, that's creating jobs, that's technology transfer. And we would exclude that from our local content metric? Person #21: The policy is clear. The policy says "majority-owned by UAE nationals". If we broaden it, we dilute the measure. And the entities that have been working on meeting their targets—they've been buying from national suppliers. If we now say "anyone with a factory counts", they'll say the rules changed mid-game. Person #6: If I may, I think there's a middle path. In the UAE, we have the national in-country value program, which uses a different scoring system. They have a points-based approach. Maybe we can adopt something like that. But that is a longer conversation, and I know the planning calendar is tight. Person #5: Is it? I mean, in the long term, yes. But for this quarter, we just need a definition. I think the registered approach is the safer one, to be honest. The data is there, it's verifiable, and the entities already know how to report it. Person #3: So we agree on the local content definition? Use the Standards? Person #21: Yes, from my side. Person #5: Yeah, I mean, if the data is clean, let's go with it. But I want it noted that the digital strategy team thinks this metric undercounts our actual local economic contribution. Person #2: Noted. What's the third one? Person #3: Entity compliance rate. The KPI is "Percentage of government entities compliant with DGE standards and mandates". The issue is what we mean by "compliant". Finance has proposed that compliance means no unresolved non-compliance findings from the last three audits. The planning template says compliance means all standards have been acknowledged and the entity has submitted its quarterly self-assessment. Two very different things. Person #21: Yes. The audit-based approach is the only credible one. Self-assessment—I mean, the entities are going to say they're compliant. It's human nature. The audits are independent, they're conducted by my team on a rotation basis, and they produce a clear finding. No finding, compliant. Finding, not compliant. Person #5: But the audit cycle is—what, every two years? Three? If an entity was audited in 2024 and had no findings, and then in 2025 they did something wrong, we wouldn't know until the next audit. So the compliance rate would be falsely high. Person #6: And also, the self-assessment has a different purpose. It's not about policing. It's about capturing where the entities are struggling, so we can provide support. If we switch to audit-only, we lose that early warning signal. The entities are more honest in the self-assessment when they know it's not a punishment—well, some of them, anyway. Person #21: The self-assessment response rate last quarter was sixty-two percent. That's not a compliance rate, that's a willingness-to-fill-forms rate. I don't see how we can present that to ADEC as "entity compliance". Person #3: For the record, the self-assessment rate was sixty-two percent because the deadline was moved to align with the Ramadan calendar, and two entities had data issues. We got a hundred percent by the fifteenth of July, but the reporting cut-off was the thirtieth of June. So it's a cut-off issue, not a compliance issue. Person #2: So what you're saying is, if we extend the cut-off, the self-assessment rate goes up. Person #3: Yes. But the ADEC reporting calendar is fixed. They want the data by the end of the quarter. We can't move the quarter. Person #5: Can we combine? Have the compliance rate be the audit-based number, and then—kind of like what ya akhi said—have a supplementary indicator on self-assessment engagement? Separate metrics, same scorecard. Person #2: That's what we said for savings. So maybe the pattern is: the main KPI is the verifiable one, and the supplementary one tells the story. Does anyone disagree with that approach in principle? Person #6: It's a good approach. It gives us the rigor and the narrative. When we did this in Riyadh, we found that the supplementary indicators often mattered more for the leadership conversations than the main ones. Person #21: I'm fine with that, as long as the main indicators are the ones defined in the Standards. And the supplementary ones are clearly labeled as "explanatory". Person #3: Okay. So we have a pattern. But I need to flag: the planning template has to be updated, and it was already distributed. If we change the definitions, the entities need to be informed. And they need to report against the new definitions for this quarter, which is—yaani—less than two weeks away. Person #5: So basically, emergency comms? We send a circular to all fifty-three entities, we say "new definitions, here's the KPI manual, here's the annex, please use this for Q3". And we just have to trust that they read it. Person #3: They don't read the circulars. We know this. Person #2: Then we make the definitions foolproof. We put them in the submission template itself, with an example calculation. No annex, no manual. The definition is in the cell. Person #3: That's what we did last year for the headcount KPI, and it reduced the queries significantly. Person #5: Yeah, that was a good fix. Actually, the headcount one—did we use the in-template definition or did they still call? Person #3: We had four queries. Previously we had thirty. Person #6: Mafeesh mushkila. Four is fine. Thirty is not. So in-template definitions it is. Person #2: Okay, so we have a pattern for the first three. And I think we can present a recommendation to the Undersecretary: main KPIs per the Standards, supplementary indicators for context, definitions embedded in the template. Now, the digital adoption one. What's the issue there? Person #5: So basically, the KPI is "Percentage of active users on TAMM". The denominator. We defined it as the number of civil servants in the Abu Dhabi government. But ADEC's feedback was that civil servants are not the only users. TAMM serves the public. So the denominator should be the eligible population of Abu Dhabi—or at least the digital-eligible population. Person #3: Wait, we discussed this in the planning workshop. And I recall we agreed that, because the KPI was in the "internal digital enablement" section of the scorecard, the denominator would be civil servants. Did that not carry through to the submission? Person #5: It did, but ADEC said no. They saw "TAMM" and they thought it was a public-facing KPI. So they didn't even look at the section header. They just—you know—they bounced it. Person #6: This happens. When we did the e-Health in Riyadh, we had a KPI that was in the internal section, but the ministry saw it as a public one. The fix was we renamed it. We added "internal" to the name. So it was "Internal TAMM adoption". And the denominator was clear from the name. Person #2: That's a simple fix. But is it the right fix? Let me ask: what is the actual thing we're trying to measure? Is it whether civil servants are using TAMM for their internal processes? Or is it whether the population is using TAMM for government services? Person #5: The original intent—from the digital strategy—was about the population. We have a vision of AI-native government, and TAMM 4.0 is the front door. The civil servant usage is a means to an end. The end is citizen adoption. Person #21: But the scorecard section is "Internal Government Efficiency". If we put a citizen-facing metric there, it's inconsistent. Person #5: Which is why it was bounced. Person #2: So we have a labeling problem. The KPI is in the wrong section. Or it's the right section but the wrong name. Or—and this is the one I'm interested in—we have the wrong KPI. What is the KPI you want, head of digital strategy? Person #5: I want "Percentage of Abu Dhabi population using TAMM for at least one government service per quarter". And I want it in the external section of the scorecard. Person #3: That's not how the scorecard works. The sections are fixed. We can't move KPIs between sections without ADEC approval. Person #5: Then we ask for approval. The Q3 submission is in two weeks. The Q4 submission is in November. We can fix it properly for Q4, and for Q3 we just use the civil servant denominator and add a footnote. Person #2: Will ADEC accept a footnote? Person #3: For one quarter, they might. They understand transitions. But they won't accept it twice. Person #5: Then we need to be clear in the cover note: "This KPI will be redefined in Q4 to align with the public adoption metric, as per the Digital Strategy 2025-2027." And we need to get the approval from ADEC's planning secretariat before the Q3 submission. That's on us. Person #2: How long does that approval take? Person #3: The secretariat meets every two weeks. If we submit the request by Thursday, it could be approved at the meeting next week. It's possible. Person #6: And if they say no, we have the footnote. So it's a hedge. Good. Person #5: So basically, we're saying: do the Q3 submission with the old definition, request the change, and if it's approved, great. If not, we accept the footnote. Person #2: I'm comfortable with that. Does anyone have a concern? Person #21: Not from Finance. The digital adoption KPI doesn't directly impact procurement. Person #2: Okay. Moving on. Are there any other definitional issues, or was it just these four? Person #3: Those are the four. But there is a non-definitional issue—the data quality for the ERP adoption KPI. The rollout to the fifty-three entities is behind schedule. We were supposed to have forty entities cut over by now. We have thirty-two. Person #5: Yeah, the Oracle Fusion rollout. That's not a definition issue, that's a delivery issue. The GovDigital team knows. They've been chasing the last few entities. Person #2: Do we need to flag it in the Q3 submission? Person #3: If we don't, ADEC will ask in the quarterly review meeting, and we'll have to explain why we didn't flag it. Person #2: Then we flag it. But we frame it as a sequencing issue, not a failure. The remaining entities are the complex ones—the ones with high transaction volumes. The pilot entities are stable. Person #5: Also, the ServiceNow rollout—the HR shared services—is on time. So we have a mixed picture. One success, one delay. Person #6: Actually, I have a question on the TAMM side. The outage in July—the one that was in the news—are we reporting anything about that in this quarter? The media coverage was not great. Person #2: That's a communications issue, not a scorecard issue. The Undersecretary's office handled the media. We don't need to report an outage on the scorecard. But if ADEC asks, we have the root cause analysis and the remediation plan. Person #5: The app rating though—the TAMM iOS rating dropped to 2.9 after the outage. Is that on the scorecard? Person #3: No. The app rating is not a scorecard KPI. It's a CX dashboard indicator, tracked separately. Person #5: But the CX team will talk about it. And the Chairman might bring it up in the Executive Council meeting. Person #2: If the Chairman brings it up, we'll handle it with the facts. The outage was four hours, it was a cloud infrastructure issue, and the remediation is in place. The rating will recover. Person #6: Insha'Allah. Person #5: Okay. And on the AI use cases—the 200+ use cases in the digital strategy—how many are actually marked "in production"? Because that was supposed to be another KPI. Person #3: The "AI use cases in production" KPI. We have—let me check my notes. I think it's forty-eight as of the end of July. The target for the end of the year is eighty. Person #5: So we're behind. But that's also a definitional issue. What counts as "in production"? Does a pilot count? Does a chatbot that's live but not fully integrated count? Person #3: The definition we have is: "AI use case is deployed in a production environment, available to end users, and monitored for performance." That's the standard. Person #5: But I know some entities have use cases that are live but not monitored. So they're not counting them. We're undercounting. Person #2: Are you saying we should loosen the definition? Person #5: I'm saying we should be clear about what counts. If it's live and available, it counts, even if the monitoring dashboard isn't fully set up. The monitoring is a quality standard, not a precondition for counting. Person #3: With respect, that creates a gray area. If we let entities count use cases without monitoring, they'll claim everything is "in production". Person #5: But we can verify. The GovDigital team can check. They know which use cases are live. Person #2: Let me suggest a compromise. For Q3, we keep the strict definition. For Q4, we work with GovDigital to create a verification process that allows us to count use cases that are live and available, with a note on monitoring status. We present both numbers: "in production" and "in production with monitoring". That gives us the full picture. Person #5: Fine. I'll take "in production with monitoring" as a separate number. But I want it in the submission. Not just in an appendix. In the main body. Person #3: The main body has a limit of three pages. We have eight KPIs, each with definition, target, actual, variance, and commentary. Adding a second number for one KPI would push us over. Person #2: We can make it fit. The CX team has commentary on the TAMM rating that can be trimmed. I'll have a look at the layout. Person #5: Also, while we're on AI—the Copilot rollout to 35,000 civil servants. That's a number I see floating around, but no one has set a KPI for it. Are we tracking it? Person #3: It's not in the scorecard. It's in the GovDigital dashboard, but not the ADEC scorecard. The Chairman's office didn't want it as a formal KPI. They said it's an enabler, not an outcome. Person #5: Okay. Then I'll stop bringing it up. But if the Chairman asks about it in the Executive Council, I want us to be ready with a number. Person #6: It's roughly twenty-one thousand licenses activated as of last week. I saw a report from the digital team. Person #5: Wait, actually—twenty-one thousand? That's—sixty percent. That's better than I thought. I was going to say I think it's around fifteen. Person #6: No, twenty-one. They distributed the licenses in May, and the activation is going smoothly. The usage is the issue. Only about forty percent of the activated licenses are used weekly. Person #2: That's a different conversation. Let's not open it now. Person #3: So, to summarize what we have so far: procurement savings definition—decision to be made after we hear the remaining items. Local content—agreed on the Standards definition. Entity compliance—agreed on audit-based main indicator with self-assessment supplementary. Digital adoption—agreed to request a section move for Q4, use current definition for Q3 with a footnote. AI use cases—agreed to keep strict definition for Q3, explore a verification process for Q4. Is that accurate? Person #5: Yes. And the ERP rollout—we flag it as a sequencing issue. Person #3: Yes. And the digital adoption—the request to move sections—that goes to the ADEC secretariat. I'll draft the request today. The Q3 submission template needs to be updated with the in-template definitions. I'll do that by Wednesday. Person #2: And the savings decision—we need a recommendation. What do we recommend? Person #21: My recommendation is to use the Standards definition. The savings number will be lower, but it will be verifiable. And we can add the administrative savings as a supplementary indicator, with a note that it's based on estimates. Person #5: But if the number is lower, the variance will be bigger, and it will look worse. Person #21: The variance is what it is. I'd rather have a lower, verifiable number than a higher, unverifiable one. When the auditors come, they check the numbers. If they can't verify, they'll flag it. Then we have a problem in front of ADEC. Person #6: I agree with the director. In Riyadh, the unverified numbers caused bigger problems than the low numbers. The trust is more important. Person #2: Does anyone disagree? Person #3: With respect, I have a concern. The target of seven hundred and fifty million was set by the Undersecretary. If we report a lower number, the Undersecretary will have to explain it to the Executive Council. He may not be happy. Person #21: I've had that conversation with him. He understands the difference between the target and the definition. He said he wants verifiable numbers. Person #3: Oh. Then that changes things. If the Undersecretary has indicated a preference for verifiable numbers, then we go with the Standards definition. Person #2: Good. So we have a recommendation. I'll present it to the Undersecretary this afternoon for final sign-off. Now, and I'm sorry to add this at the end, but the Chairman's office asked about the Tomouh learning platform completion rate. It's a small thing, but it's on their mind. Person #3: The Tomouh completion rate—it's in the Government Talent sector's dashboard, not in the ADEC scorecard. But we received a query. I have the numbers. The overall completion rate is thirty-eight percent. The target is sixty by end of year. Person #5: Wait, Tomouh? The learning hub? Only thirty-eight percent? I thought we were doing better than that. Person #3: The completion rate is low because many courses are long-form. The recent push for microlearning modules has been more successful—those have a seventy percent completion rate. But the overall number is dragged down by the long-form courses. Person #2: The Chairman's office wants a better story. Can we create a narrative around microlearning? Person #3: We can. But I need the Government Talent team to provide the microlearning numbers. We'll follow up. Person #2: Okay. Let's make that a follow-up item. We'll need it before the end of the week. Person #5: Quick question, actually. The 2026 HR Law—the meritocracy thing—are we tracking any of that on the scorecard? I know it's not a KPI, but is there a plan for next year? Person #3: Not in the current scorecard. Government Talent is doing a gap analysis. It's a major policy change, so they're not rushing the KPIs. It will probably be part of the 2027 planning cycle. Person #5: Okay. That makes sense. You can't measure something that's being rolled out. Person #2: Exactly. Let's not get ahead of the policy implementation. Person #6: If I may ask about the World Governments Summit—is that in the planning cycle? Because the February date is coming, and the preparations take time. Person #2: That's a Corporate Services function. They have their own timeline. It's not a scorecard issue. Person #6: I know, but the Chairman might want to highlight the digital strategy on that stage. If we have a strong number by then, it's a good story. Person #5: That's actually a great point. By February, we might have the AI use cases number looking good. And the TAMM 4.0 adoption—if we fix the definition, we can present a compelling case. Person #3: The submission for the summit to include the digital strategy is due to the summit office in September. We'll need a highlight reel. But that's for later. Let's not open that thread now. Person #2: Right. We're getting to the end of our time, and I want to make sure we have clear action items. Can we do a quick recap? Person #30: Just to recap—and I have this in my notes—the action items from today: one, update the Q3 submission template with the agreed definitions, embedded in the cells, owner is the planning team. Two, draft the request to ADEC secretariat for the digital adoption KPI section move, owner is planning team. Three, prepare the savings recommendation and present to the Undersecretary, owner is the DG. Four, follow up with GovDigital on the AI use case verification process. Five, follow up with Government Talent on the Tomouh microlearning numbers. Did I miss anything? Person #3: The local content and ERP rollout items. And the entity compliance self-assessment supplementary indicator. Person #30: Yes, so: six, prepare the local content definition memo for the entities. Seven, prepare the ERP rollout flag for the Q3 submission. Eight, coordinate with GovDigital on the entity compliance self-assessment supplementary indicator. I'll send the notes by the end of the day. Person #2: Good. And Sameera, can you share the template with the planning team for the in-cell definitions? Person #20: Yes, I'll send it today. And can I ask—the local content memo—should we include the example calculation? Like the one we used for the headcount KPI? Person #3: Yes, please. The example calculation is essential. If there's any ambiguity, the entities will call. And we don't have time for calls. Person #20: Understood. Noted. Person #6: Ya basha, one more thing. The TAMM program board meets next week. Should I mention the digital adoption KPI change to them? They're sensitive about any changes to the service levels. Person #2: It's not a service level change. It's a reporting definition. But you can brief them informally. Don't make it a formal item. Person #6: Okay. I'll have a coffee with the program director before the meeting. He owes me one. Person #5: About the AI use case verification—should we also involve the GovDigital data team? They have the monitoring dashboards. They could provide the "in production with monitoring" numbers. Person #3: Yes. That would help. But can we do it in the next two weeks? The submission is on the fifteenth. Person #5: I'll talk to them today. If we can get the numbers by next week, we can decide whether to include them in the Q3 submission or wait. If it's tight, we wait. Person #3: That's fine. The Q3 submission can have the strict number, and the Q4 submission can have the fuller picture. Person #2: Good. And one last thing. The Chairman asked for a one-page summary of the quarterly performance for his own reading before the ADEC meeting. He wants it by the tenth. The planning team will prepare it, and the sectors will contribute their commentary. I'll circulate the format today. Person #3: Thank you. We'll work on it. Person #2: Okay. I think we're good. Let me know if anything else comes up in the meantime. Person #5: Can I just—one question about the TAMM hallucinated fees incident? I know it's not on the scorecard, but are we tracking the complaint volume? Person #3: We are tracking it, yes. The complaint volume is monitored by Customer Experience Affairs, not through the scorecard. But if there is a spike, it gets flagged in the weekly risk note to the Undersecretary. Person #5: Right, but the issue last week—the fee hallucination on the TAMM assistant—that was pretty public. I saw people posting screenshots. Do we have a number? Person #6: Yaani, we saw about three hundred and twenty complaints in the first forty-eight hours. Most of them were "the assistant told me my parking fine was one thousand dirhams" when it was actually thirty. Some people actually paid, which is the worrying part. Person #21: That's a service quality issue, not a scorecard issue. Under the Standards for Government Procurement, clause 4.2, we don't measure operational incidents. We measure outcomes. Person #5: I get that, but the Chairman's office asked about it last week. They wanted to know if we were going to include a customer trust indicator in the Q3 submission. Person #2: Let me frame this. The complaint volume is an operational metric. For the ADEC scorecard, we need a stable set of indicators. If we start adding incident-based metrics every time something goes wrong, we'll be chasing the news cycle. The Digital Strategy has a KPI on digital adoption, and we have a separate CX outcome measure. We can discuss whether to feed the complaint data into the narrative, but not as a separate KPI. Person #3: Correct. And for the record, the Q3 submission is due on the fifteenth, so if we were to add a new definition, we would need the standards committee to approve it. That's not realistic. We can mention it in the commentary. Person #20: Noted. So we keep it as a narrative point. Should I ask Customer Experience Affairs for a clean summary? I can get it by Thursday. Person #2: Yes, please. Keep it short. One paragraph, and focus on the remediation, not the incident. Person #5: And that's already happening, right? The TAMM team said they're retraining the model on the fee schedules. We just need the language to say "the assistant was updated immediately. Person #6: Ya basha, I have a call with them this afternoon. I can confirm the retraining is done. The problem is the knowledge base had an outdated tariff table from before the merger. That's the kind of thing we keep flagging about the data governance. Person #3: And that's a good point to include in the data governance section of the Q3 submission. But let's not conflate it with the scorecard. Person #21: Agreed. And while we're on the subject of definitions—the savings target. The Undersecretary asked me to confirm that we're still using the baseline from the 2024 procurement audit. I have some entities that are pushing back, saying they want to include in-kind contributions. Person #3: For the scorecard, we use the approved definition. Savings are cashable savings under framework agreements. In-kind contributions are not counted. That's in the standards library, version 3.1. Person #21: Yes, I know. But the Department of Culture and Tourism is arguing that their co-hosting of an event with a supplier should count as a contribution. I told them no, but they want a formal interpretation. Person #2: We can handle that through the standards committee. Do you have their request in writing? Person #21: I do. I'll circulate it after the call. Person #20: And on that, the new framework for IT hardware is getting a lot of interest. We have forty-two suppliers registered on the Procurement Gate. The award is scheduled for mid-September. Person #3: Good. That's a concrete deliverable we can note. But just to be clear, the framework award itself is not a scorecard KPI. The savings realization from the framework utilization is what we measure. Person #20: Understood. As soon as the first purchase orders come through, we'll map them to the savings tracker. Person #6: Which reminds me—when we did this in Riyadh, we had a similar issue. The government procurement office created a new framework, but nobody used it for the first two months because the entities were still on their legacy contracts. So the savings were zero. We had to create a transition period and then measure the adoption rate. Person #5: So basically, you're saying we should track the utilization rate separately from the savings? Person #6: Exactly. Utilization, then savings. But that's for the sector to decide, not for the scorecard. Person #3: For the scorecard, we keep the savings target as defined. We can note the utilization in the commentary. Person #2: Good. And that's the principle for this whole review. The scorecard is for outcomes, not activities. Every time we add an activity metric, we dilute the signal. Person #5: Okay, I hear you. But can I push back a little? Sometimes the activities are what we can actually influence. Like the AI use case count—we have a target of two hundred use cases by the end of the strategy, and we're at one hundred and thirty-seven. That's a count, not an outcome. But the outcome is hard to measure. So we count the use cases and then, in the Q4, we'll do the impact assessment. Person #2: That's a fair point. The Digital Strategy has its own KPI set, and we've agreed to treat it as a separate annex to the scorecard. So the AI use case count can stay there, with the verification that you and the GovDigital team are working on. Person #3: Yes. And for the Q3 submission, the planning team will only include the verified number from the monitoring dashboards. If that takes more than two weeks, we wait. Person #5: I'm on it. I'll call the GovDigital data team today. If they can give me the numbers by next Thursday, we can verify. Honestly, they're a little protective of their dashboards, but I think they'll play ball. Person #2: Good. Now, can we go back to the Chairman's one-page summary? I said I'd circulate the format today. The planning team will prepare the draft, but I need each of you to send me your commentary points by Wednesday. Not paragraphs—bullet points. Three to five bullets. Person #21: For Finance, my points are: savings target on track, framework utilization improving, and one risk—entity non-compliance with the procurement standards. Person #20: And the new IT hardware framework. That's a positive item. Person #21: Yes, but that's not confirmed yet. We can say "award expected in September" and then adjust in the final version. Person #5: For digital strategy: AI use cases validation in progress, digital adoption steady at ninety-one percent, and the TAMM assistant incident as a risk with remediation completed. Person #2: Good. And for the CX side, we'll need something from them. I'll follow up with the Customer Experience Affairs Director directly, or maybe one of you can. Person #3: I'll draft a request to them this afternoon. I know they're busy with the CX Awards prep, but a few bullet points shouldn't take too long. Person #6: Yaani, if you want, I can call the CX director. We go back a long time. I can explain the context and get her to keep it short. Person #2: That would be helpful. Just don't let her expand into a full report. Person #6: Mafeesh mushkila. I'll say "three bullets, no more". Person #30: Just to recap the action items so far: one, I'll request the TAMM complaint summary from CX; two, the AI verification with GovDigital by next Thursday; three, each sector sends the DG their bullets by Wednesday; four, the standards committee will circulate the interpretation request from DCT. Did I miss anything? Person #3: No, that's correct. And the Chairman's summary draft will be ready by the tenth. Person #20: And the procurement savings tracker update? Should I send that by Friday? Person #21: Yes, send it to me and the planning team. But wait until the end of the week so we have the latest purchase orders. Person #20: On it. Person #5: Can I ask one more thing? The Youth Council has an idea for a TAMM challenge—a hackathon-style competition for government employees to redesign a service journey. Would that be something we can include in the CX narrative for the Chairman's summary? It's a positive, forward-looking item. Person #3: For the scorecard, no. For the narrative, yes, but you need to keep it to one line. Person #2: I like the idea, but let's not overpromise. The Chairman has seen many hackathons. If you can show it's linked to a real service improvement, then it's worth mentioning. Otherwise, it might look like an activity. Person #5: No, I get that. We're actually partnering with MBZUAI and Hub71 on the challenge, and the winning team will get to pilot their journey in TAMM. So it's a real thing. Person #6: In Riyadh, we did something similar with the municipality. The winning team's chatbot went into production in three months. It was a huge success. Person #5: Exactly. So it's not just a feel-good event. Person #2: Good, then you can include one line in the narrative. But make sure it's clear that the outcome is the pilot, not the hackathon. Person #21: And if the pilot involves a procurement component, please run it through the Standards for Government Procurement. We don't want any surprises. Person #5: Of course. We'll work with your team on the framework. Person #3: Can we go back to the data governance point for a second? The TAMM incident highlighted a broader issue with the tariff tables. We have a data quality standard in the library, but it's not being applied consistently. Do we need to escalate this to the Data Council, or can GovDigital handle it? Person #6: I raised it in the last GovDigital coordination meeting. They said they're creating a new data steward role for the TAMM knowledge base. It should be filled by the end of September. Person #2: Good, but for the October GITEX, we don't want another public incident. Maybe we should have a pre-GITEX review of all the data that feeds the assistant. Person #5: That's a good idea. I'll add it to the digital strategy risk register. We can do a walkthrough of the data sources by mid-September. Person #30: I can coordinate that. I'll set up a meeting with GovDigital and the TAMM team. Person #2: Thank you. Now, is there anything else on the scorecard definitions? I know we have a few pending items from the entities. Person #3: Yes, there are two. The first is from the Department of Education and Knowledge—they want to add a "student satisfaction" indicator to the scorecard, but it's not in the approved standards library. We told them we'd review it after the Q3 submission. Person #21: That's a new metric, not a definition change. Under the Standards, any new KPI needs to go through the full cycle. So Q4 at the earliest. Person #3: Correct. The second is from the Department of Municipalities and Transport—they want to change the definition of "service completion rate" to exclude transactions that are withdrawn by the customer. They say the current definition penalizes them for transactions that were aborted. Person #2: That's a legitimate definition question. What does the current standard say? Person #3: The standard says "transactions completed as a percentage of transactions initiated." Withdrawn transactions are counted as initiated. So they fall into the denominator. Person #6: Obviously, that's not fair. If the customer withdraws the transaction, the entity had no control. In Riyadh, we had a similar case and we changed the definition to exclude withdrawals after a certain threshold. Person #5: Maybe we can add a sub-category: "completed, pending, withdrawn" and report the completion rate only on the completed ones. But that would change the KPI. Person #2: Let's not change it now. We can ask the standards committee to interpret it, and for the Q3 submission, we keep the current definition. If they want to appeal, they can do so after Q3. Person #3: With respect, if we keep the current definition, they'll claim the Q3 numbers are misleading. And they might escalate to the Undersecretary. Person #2: Then we should get ahead of it. Let's ask their entity to submit a formal request for a definition review, and we'll add it to the Q4 planning cycle. But the Q3 numbers will be reported under the current definition, with a note that a review is pending. Person #3: That's reasonable. I'll draft the note for the commentary. Person #21: And just to be clear, the same principle applies to procurement—we don't change definitions mid-cycle. Person #5: Okay, so no changes mid-cycle. Got it. Person #2: Anything else? We have about ten minutes left, and I need to prepare for another meeting. Person #30: Just to confirm, the next scorecard working group is scheduled for the twenty-first of September, right? Person #3: Yes, the twenty-first at ten in the morning. This room. Person #30: Perfect. I'll send the calendar invite. Person #6: And I'll have the TAMM program board update by then. If they raise any concerns about the digital adoption KPI, I'll bring them to the working group. Person #2: Good. And please, when you speak to the program director, make sure he understands it's a reporting definition, not a change to the service level. We don't want another committee discussion about it. Person #6: Yaani, I'll say it's a definitional alignment, khalas. He's a reasonable man, as long as it's not a formal item. Person #5: Do we have a sense of the total number of Q3 submissions? Are all fifty-three entities going to submit on time? Person #3: We expect forty-nine to submit on time. Four have already requested extensions—two from the Department of Health cluster and two from the municipality cluster. We'll manage the exceptions. Person #20: The Department of Health ones, are those because of the system migration to the new ERP? Person #3: Yes, they're in the middle of the Oracle Fusion rollout, so their finance data is a little delayed. We're working with them to get the performance data through a manual upload. Person #21: I was told the Oracle rollout is behind schedule. Is that still on track for the end of the year? Person #5: It's on track, but it's tight. GovDigital is supporting them with extra data migration resources. The go-live is still scheduled for the first of December, before National Day. Person #6: In Riyadh, the big ERP rollout was delayed by three months because of data migration. Let's hope we don't have the same issue. Person #21: If the rollout is delayed, the procurement module is delayed, which means the savings tracker will get more complicated. I hope they meet the date. Person #5: So do I. But the oracle team has been pretty transparent. They're doing a dress rehearsal in October, and they'll do a parallel run in November. Person #2: Okay, that's a Q4 issue, not a Q3 issue. But keep me informed if the risk escalates. Person #5: Will do. Person #20: Just to add, I saw the draft press release for GITEX. Is the new TAMM 4.0 feature going to be announced there? We might need to update some procurement materials. Person #5: Yes, TAMM 4.0 is being beta-tested now. The launch is at GITEX. But I don't think we need a procurement update for that. Person #20: No, but we're working on a smart procurement journey on TAMM for suppliers—electronic bids, e-contracts, and so on. It's part of the procurement modernization. We want to announce a pilot at GITEX. Person #21: That's been in the plan for a while. We need to be careful not to overpromise. The pilot has to be real, not a mock-up. Person #20: It's real. We have three suppliers lined up, and the development is almost complete. We'll show a live demo, inshallah. Person #2: That's a good thing to include in the meetings with delegations. The Chairman likes to show concrete progress at GITEX. Person #6: We have a delegation from Singapore visiting in early October. I can mention the smart procurement journey as a new initiative. Person #2: Yes, just coordinate with the protocol team so we don't double-book. And make sure the demo is working the day before. Person #20: Yes, we're doing a full rehearsal. No surprises. Person #3: For the Q3 scorecard submission, should I add a placeholder for the smart procurement pilot? Or is it too early? Person #21: Better to keep it in the commentary as a "forward-looking initiative." The pilot isn't a KPI until it's measured. Person #3: Agreed. I'll add a placeholder in the narrative but not in the dashboard. Person #2: Anything else from anyone? I have a call at eleven. Person #30: Quick question—should I share the minutes of this call with the full planning team or just with the sector heads? Person #2: Share with the planning team and the sector heads. The Undersecretary can see it if he asks. Person #30: Okay. I'll circulate the action items by end of day. Person #5: And I'll send the AI verification request to GovDigital right after this. If I get a response, I'll update the thread. Person #3: Thank you. And I'll send the commentary template to you all before noon. Person #6: I'll call the CX director this afternoon and make sure she sends the three bullets. Person #21: And I'll circulate the DCT interpretation request to the standards committee for the record. Person #2: Good. I think we're aligned. Let's keep the momentum, and if anything urgent comes up, you have my direct line. Person #30: Just to recap, the key dates—submission on the fifteenth, Chairman's summary by the tenth, and the working group on the twenty-first. Person #2: Right. Anything else? Person #5: No, I think that's it. Person #6: Ya basha, mafeesh mushkila. Everything will be ready. Person #20: Noted, thank you everyone. Person #3: Thank you, everyone. We'll speak soon. Person #2: Shukran, khalas. Bye.