vidtheque.

read-only demo

Videos jt1Pbr_n6oU

Your Moat Is Your Data Model — Mike Phipps, Gates Foundation

index_state ready data_status ok

AI Engineer· published 2026-07-22· 0:20:29· en-US· indexed 2026-08-10 19:44

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:03, 1 of 1 keyframes kept
  2. Shot 1, 0:03 to 0:05, 1 of 1 keyframes kept
  3. Shot 2, 0:05 to 0:12, 1 of 1 keyframes kept
  4. Shot 3, 0:12 to 0:25, 1 of 1 keyframes kept
  5. Shot 4, 0:25 to 0:50, 1 of 1 keyframes kept
  6. Shot 5, 0:50 to 1:16, 1 of 1 keyframes kept
  7. Shot 6, 1:16 to 1:42, 0 of 1 keyframes kept
  8. Shot 7, 1:42 to 2:08, 1 of 1 keyframes kept
  9. Shot 8, 2:08 to 2:34, 0 of 1 keyframes kept
  10. Shot 9, 2:34 to 2:50, 1 of 1 keyframes kept
  11. Shot 10, 2:50 to 3:20, 1 of 1 keyframes kept
  12. Shot 11, 3:20 to 3:50, 0 of 1 keyframes kept
  13. Shot 12, 3:50 to 4:37, 1 of 1 keyframes kept
  14. Shot 13, 4:37 to 5:03, 1 of 1 keyframes kept
  15. Shot 14, 5:03 to 5:22, 0 of 1 keyframes kept
  16. Shot 15, 5:22 to 5:47, 0 of 1 keyframes kept
  17. Shot 16, 5:47 to 6:12, 1 of 1 keyframes kept
  18. Shot 17, 6:12 to 6:40, 0 of 1 keyframes kept
  19. Shot 18, 6:40 to 7:07, 1 of 1 keyframes kept
  20. Shot 19, 7:07 to 7:44, 1 of 1 keyframes kept
  21. Shot 20, 7:44 to 8:21, 0 of 1 keyframes kept
  22. Shot 21, 8:21 to 8:52, 0 of 1 keyframes kept
  23. Shot 22, 8:52 to 9:19, 1 of 1 keyframes kept
  24. Shot 23, 9:19 to 9:46, 1 of 1 keyframes kept
  25. Shot 24, 9:46 to 10:13, 0 of 1 keyframes kept
  26. Shot 25, 10:13 to 10:39, 1 of 1 keyframes kept
  27. Shot 26, 10:39 to 11:06, 1 of 1 keyframes kept
  28. Shot 27, 11:06 to 11:33, 1 of 1 keyframes kept
  29. Shot 28, 11:33 to 12:00, 0 of 1 keyframes kept
  30. Shot 29, 12:00 to 12:27, 0 of 1 keyframes kept
  31. Shot 30, 12:27 to 12:54, 0 of 1 keyframes kept
  32. Shot 31, 12:54 to 13:20, 1 of 1 keyframes kept
  33. Shot 32, 13:20 to 13:47, 1 of 1 keyframes kept
  34. Shot 33, 13:47 to 14:19, 1 of 1 keyframes kept
  35. Shot 34, 14:19 to 14:52, 1 of 1 keyframes kept
  36. Shot 35, 14:52 to 15:24, 0 of 1 keyframes kept
  37. Shot 36, 15:24 to 15:53, 1 of 1 keyframes kept
  38. Shot 37, 15:53 to 16:22, 0 of 1 keyframes kept
  39. Shot 38, 16:22 to 16:51, 1 of 1 keyframes kept
  40. Shot 39, 16:51 to 17:20, 0 of 1 keyframes kept
  41. Shot 40, 17:20 to 17:50, 0 of 1 keyframes kept
  42. Shot 41, 17:50 to 18:19, 1 of 1 keyframes kept
  43. Shot 42, 18:19 to 18:57, 1 of 1 keyframes kept
  44. Shot 43, 18:57 to 19:25, 1 of 1 keyframes kept
  45. Shot 44, 19:25 to 19:56, 0 of 1 keyframes kept
  46. Shot 45, 19:56 to 20:03, 1 of 1 keyframes kept
  47. Shot 46, 20:03 to 20:12, 1 of 1 keyframes kept
  48. Shot 47, 20:12 to 20:29, 0 of 1 keyframes kept

48 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
229
whisperx 229
chunks
35
from 229 cues
keyframes
30
kept of 48 captured
frames with text
30
674 lines read
chapters
11
from the source metadata
keyframe bytes
5.6 MB
word timings on 229 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-10 00:33 1m 51s
stt done — 2026-08-10 00:35 21s
chunk done — 2026-08-10 00:35 0s
text_embed done — 2026-08-10 19:44 1s
keyframe done — 2026-08-10 00:35 1m 58s
ocr done — 2026-08-10 00:37 15s
frame_embed done — 2026-08-10 19:44 5s

Frames, and what the machine read

  • 0:02 #0 done2 line(s)

    shot 0·sharpness 453.4

    1. AlEngineer0.96
    2. World's Fair0.97

  • 0:03 #1 done2 line(s)

    shot 1·sharpness 665.7

    1. AlEngineer0.95
    2. World's Fair0.99

  • 0:10 #2 done24 line(s)

    shot 2·sharpness 2743.2

    1. LAB & PLATINUM SPONSORS0.98
    2. Amazon AGI Lab0.98
    3. ANTHROP\C1.00
    4. Google DeepMind1.00
    5. MINIMAX0.94
    6. OpenAI0.92
    7. Akamai1.00
    8. arize1.00
    9. aws1.00
    10. Braintrust bright data0.99
    11. B1.00
    12. Browserbase1.00
    13. docker1.00
    14. :neo4j0.93
    15. ORACLE1.00
    16. PayPal1.00
    17. qodo1.00
    18. reducto1.00
    19. Sonar1.00
    20. Makers of0.99
    21. togetherai1.00
    22. Unblocked1.00
    23. WorkOS1.00
    24. SonarQube1.00

  • 0:23 #3 done13 line(s)

    shot 3·sharpness 1521.1

    1. 1/261.00
    2. AlEngineer0.99
    3. World'sFair1.00
    4. Your Data Model is Your Moat1.00
    5. PRESENTED BY1.00
    6. Strategic Intelligence Platform (SIP) Enterprise Rollout1.00
    7. Microsoft1.00
    8. Mike Phipps1.00
    9. Lead Al Engineer, Knowledge Management & Insights0.99
    10. Gates Foundation0.99
    11. ×0.90
    12. World'sFair1.00
    13. Engineering the future of Al1.00

  • 0:35 #4 done22 line(s)

    shot 4·sharpness 2481.1

    1. 2/260.99
    2. AlEngineer0.97
    3. World'sFair1.00
    4. Intro: What's defensible for your organization1.00
    5. • Pace of change of Al is extremely fast but at the same time tools like Claude Code make it possible for dev to move even faster0.99
    6. • However, in production there are constraints:0.99
    7. • how much of the deployed stack are you prepared to own?1.00
    8. PRESENTED BY1.00
    9. • how much appetite do users have for decentralized access?0.99
    10. • what's your product differentiation from SaaS platforms that are widely used?0.99
    11. Microsoft1.00
    12. · Forced reflection:0.95
    13. • what's your team's skillset and competitive advantage within that environment?0.99
    14. • For us (and what I hope you take away from this presentation!):1.00
    15. • our moat is our understanding of internal processes and the tacit knowledge needed for any Al (no matter how good!) to1.00
    16. work effectively across our data1.00
    17. • These processes and tacit knowledge are modeled within the Strategic Intelligence Platform (SIP) that we launched into production on1.00
    18. June 4 for enterprise-wide internal use at the Gates Foundation.1.00
    19. 2 luly 20260.82
    20. At tnginee Word's Ffar0.72
    21. World'sFair1.00
    22. Engineering the future of Al0.98

  • 1:11 #5 done23 line(s)

    shot 5·sharpness 2337.2

    1. 2/261.00
    2. AlEngineer0.97
    3. World'sFair1.00
    4. Intro: What's defensible for your organization1.00
    5. • Pace of change of Al is extremely fast but at the same time tools like Claude Code make it possible for dev to move even faster0.99
    6. • However, in production there are constraints:0.99
    7. • how much of the deployed stack are you prepared to own?0.99
    8. PRESENTED BY1.00
    9. • how much appetite do users have for decentralized access?0.99
    10. • what's your product differentiation from SaaS platforms that are widely used?0.99
    11. Microsoft1.00
    12. · Forced reflection:0.95
    13. • what's your team's skillset and competitive advantage within that environment?0.99
    14. • For us (and what I hope you take away from this presentation!):0.98
    15. • our moat is our understanding of internal processes and the tacit knowledge needed for any Al (no matter how good!) to1.00
    16. work effectively across our data1.00
    17. • These processes and tacit knowledge are modeled within the Strategic Intelligence Platform (SIP) that we launched into production on0.99
    18. June 4 for enterprise-wide internal use at the Gates Foundation.1.00
    19. 2 ludy 20260.86
    20. Al Engieer ord's Fair0.82
    21. World'sFair1.00
    22. TRACK 5· JULY 2, 20260.94
    23. Graphs1.00

  • 1:27 #6 skipped

    shot 6·duplicate of #5

  • 1:45 #7 done21 line(s)

    shot 7·sharpness 2198.7

    1. 2/261.00
    2. AlEngineer0.96
    3. World'sFair1.00
    4. Intro: What's defensible for your organization1.00
    5. • Pace of change of Al is extremely fast but at the same time tools like Claude Code make it possible for dev to move even faster0.99
    6. • However, in production there are constraints:0.99
    7. • how much of the deployed stack are you prepared to own?1.00
    8. • how much appetite do users have for decentralized access?1.00
    9. • what's your product differentiation from SaaS platforms that are widely used?0.99
    10. · Forced reflection:0.95
    11. • what's your team's skillset and competitive advantage within that environment?0.99
    12. • For us (and what I hope you take away from this presentation!):0.99
    13. • our moat is our understanding of internal processes and the tacit knowledge needed for any Al (no matter how good!) to1.00
    14. work effectively across our data1.00
    15. • These processes and tacit knowledge are modeled within the Strategic Intelligence Platform (SIP) that we launched into production on0.99
    16. June 4 for enterprise-wide internal use at the Gates Foundation.0.99
    17. 2 ludy 260.80
    18. Al Engieer ord's Fair0.84
    19. World'sFair1.00
    20. TRACK 5· JULY 2,20260.96
    21. Graphs1.00

  • 2:23 #8 skipped

    shot 8·duplicate of #7

  • 2:43 #9 done8 line(s)

    shot 9·sharpness 680.1

    1. 3/261.00
    2. AlEngineer0.97
    3. World's Fair0.96
    4. The Gates Foundation:1.00
    5. Overview1.00
    6. World'sFair1.00
    7. TRACK 5· JULY 2, 20260.95
    8. Graphs1.00

  • 2:56 #10 done28 line(s)

    shot 10·sharpness 1902.7

    1. 4/261.00
    2. AlEngineer0.98
    3. World's Fair0.96
    4. Our role1.00
    5. We work with partners worldwide to tackle critical problems1.00
    6. Creating market incentives for1.00
    7. Spurring innovation that improves1.00
    8. Strengthening collaboration1.00
    9. Generating high-quality data1.00
    10. life-saving products by supporting1.00
    11. the human condition by stepping1.00
    12. by bringing together the public,0.99
    13. and evidence that drives progress1.00
    14. the development and delivery of0.99
    15. in where governments can't, and1.00
    16. private and philanthropic sectors1.00
    17. by showing what's working and0.99
    18. vaccines, treatments, and tools for1.00
    19. businesses won't.1.00
    20. to save and transform lives around0.98
    21. what isn't.0.97
    22. those most in need.0.98
    23. the world.1.00
    24. 2 luly 20260.98
    25. AI Engineer World's Fair0.93
    26. World's Fair0.97
    27. TRACK 5· JULY 2,20260.95
    28. Graphs1.00

  • 3:35 #11 skipped

    shot 11·duplicate of #10

  • 4:09 #12 done27 line(s)

    shot 12·sharpness 1675.5

    1. 5/261.00
    2. AlEngineer0.97
    3. World'sFair0.99
    4. Scope of our work (Stats for new grants from F.Y. 2023)0.99
    5. We work with partner organizations around the world to reduce inequity0.99
    6. 2,3961.00
    7. 1351.00
    8. 2,2381.00
    9. 441.00
    10. Total Grants1.00
    11. Countries1.00
    12. Alumni1.00
    13. Program Strategies0.98
    14. 481.00
    15. 1,4221.00
    16. $7.75B1.00
    17. 2,0261.00
    18. U.S. States1.00
    19. Grantees1.00
    20. Total Charitable Support1.00
    21. Employees1.00
    22. For the year ended December 31, 2023. Amount in U.S. doilors.0.96
    23. 2 luly 20260.92
    24. Al Engineer World's Fair0.96
    25. World'sFair1.00
    26. TRACK 5· JULY 2,20260.95
    27. Graphs1.00

  • 4:43 #13 done28 line(s)

    shot 13·sharpness 1341.8

    1. 6/260.91
    2. AlEngineer0.99
    3. World's Fair0.99
    4. Foundation funding summary0.98
    5. In 2023, the foundation provided US $7.75B* charitable support in the following areas:1.00
    6. A. Global Development1.00
    7. $2.005B0.99
    8. B. Global Health0.99
    9. $1.860B1.00
    10. C. Global Growth & Opportunity0.98
    11. $875M1.00
    12. D. Gender Equality0.97
    13. $863M0.94
    14. Total charitable support1.00
    15. $7.75B0.99
    16. E. United States Program1.00
    17. $692M0.90
    18. F. Global Policy & Advocacy0.99
    19. $328M1.00
    20. G. Other Charitable Programs0.96
    21. $204M0.96
    22. H. Operational Expenditure0.99
    23. $922M0.97
    24. 2 luly 20260.95
    25. AI Engineer World's Fair0.94
    26. World's Fair0.98
    27. TRACK 5· JULY 2,20260.94
    28. Graphs1.00

  • 5:20 #14 skipped

    shot 14·duplicate of #9

  • 5:35 #15 skipped

    shot 15·duplicate of #7

  • 5:53 #16 done33 line(s)

    shot 16·sharpness 1484.2

    1. 8/261.00
    2. AlEngineer0.99
    3. World'sFair1.00
    4. From Systems of Record to Agentic Intelligence0.99
    5. Systems of Record0.99
    6. Enterprise Data Platform0.99
    7. PRESENTED BY1.00
    8. Doc Repository1.00
    9. Strategic Intelligence Platform1.00
    10. Microsoft1.00
    11. HR Source1.00
    12. System1.00
    13. Data Lakehouse1.00
    14. Data Curation1.00
    15. Pre-Processing1.00
    16. Cross-system semantic graph layer1.00
    17. agents reason over1.00
    18. Structured Data1.00
    19. Agentic Chat0.99
    20. Investment1.00
    21. Extraction &0.99
    22. Source System1.00
    23. Enrichment1.00
    24. Funding Source1.00
    25. Unstructured Data0.98
    26. Agentic Workflows0.98
    27. System1.00
    28. Governance1.00
    29. 2 luly 20260.92
    30. AI Engineer World's Fair0.97
    31. World's Fair0.96
    32. TRACK 5· JULY 2, 20260.96
    33. Graphs1.00

  • 6:28 #17 skipped

    shot 17·duplicate of #7

  • 7:04 #18 done18 line(s)

    shot 18·sharpness 1563.2

    1. 9/261.00
    2. AlEngineer0.98
    3. World'sFair1.00
    4. Dataset Owner Engagement is Critical!1.00
    5. • Often significant amount of tacit knowledge controlled by data owners0.99
    6. • Data owner engagement critical for:0.98
    7. • interpreting full meaning behind various fields0.98
    8. • understanding the structure of the dataset as a whole0.99
    9. • joining datasets together correctly1.00
    10. • understanding limitations or systematics of the data0.99
    11. • applying appropriate safeguards and security trimmings1.00
    12. • ensuring standardized reporting conventions0.99
    13. This is the procedural understanding and tacit knowledge needed for Al to work effectively in the enterprise!0.99
    14. 2 luly 20260.83
    15. AI Engineer World's Fair0.95
    16. World'sFair1.00
    17. TRACK 5· JULY 2, 20260.93
    18. Graphs1.00

  • 7:40 #19 done25 line(s)

    shot 19·sharpness 1183.6

    1. 10/261.00
    2. AlEngineer1.00
    3. World's Fair0.99
    4. From Raw to Curated Data0.98
    5. Data Curation1.00
    6. PRE-PROCESSING1.00
    7. EXTRACTION L ENRICHMENT0.96
    8. GOVERNANCE0.97
    9. Filtering0.94
    10. Structured Field Elxtraction0.97
    11. PII Masking1.00
    12. Semantic Chunking1.00
    13. Deduplication1.00
    14. Sensitive Content Classification1.00
    15. Multimodal Elxtraction0.99
    16. Version Tagging0.97
    17. Access & Entitlement Mapping0.96
    18. Entity & Taxonomy Tagging0.94
    19. Structured Metadata1.00
    20. Different datasets run different cowbinations of these steps.0.94
    21. 2 luly 20260.90
    22. AI Engineer World's Fair0.96
    23. World's Fair0.95
    24. TRACK 5· JULY 2, 20260.93
    25. Graphs1.00

  • 8:13 #20 skipped

    shot 20·duplicate of #19

  • 8:25 #21 skipped

    shot 21·duplicate of #9

  • 9:13 #22 done31 line(s)

    shot 22·sharpness 946.0

    1. INVESTMENT1.00
    2. FUNDING1.00
    3. MEETINGS & DOCS0.99
    4. Funding1.00
    5. Meetings & Docs0.99
    6. 12/261.00
    7. AlEngineer0.98
    8. World's Fair0.99
    9. Meeting0.98
    10. Strategy1.00
    11. HAS MEETING0.96
    12. strategy team0.96
    13. HAS_DOCUMENT0.99
    14. STEP 1- THE ANNUAL RHYTHM0.96
    15. Strategy reviews set the1.00
    16. Document1.00
    17. agenda1.00
    18. board decks · narratives0.97
    19. Each year strategy teams review progress and set1.00
    20. priorities across a 5-year horizon.0.99
    21. Strategy Reviews and Annual Planning generate board0.99
    22. decks, rankings and notes.1.00
    23. Those documents live in the curated document library0.99
    24. our unstructured corpus.0.99
    25. 80+ strategy teams across the foundation.0.99
    26. What's shown here is the conceptual data model, which1.00
    27. doesn't show instantiation or cardinality.1.00
    28. ×0.60
    29. World's Fair0.99
    30. TRACK 5· JULY 2, 20260.93
    31. Graphs1.00

  • 9:42 #23 done23 line(s)

    shot 23·sharpness 1028.7

    1. Funding1.00
    2. Meetings & Docs0.97
    3. MEETINGS & DOCS1.00
    4. 13/261.00
    5. AlEngineer0.98
    6. World's Fair0.98
    7. STEP 2·FUNDING THE PLAN0.97
    8. Budgets flow through the1.00
    9. funding hierarchy0.98
    10. Annual planning finalizes each strategy team's board-1.00
    11. approved budget.0.99
    12. Money flows from a strategy to a funding team defined1.00
    13. by a 5-level funding path: Division → Strategy → Group1.00
    14. → SubGroup → Body of Work.0.99
    15. A funding team is allocated an annual budget0.99
    16. (FundsToBoW).1.00
    17. Aliases and full text indexes on team nodes to enhance0.99
    18. agentic retrieval.1.00
    19. Funding path is a 5 level additive DAG.0.98
    20. ×0.81
    21. World's Fair0.95
    22. TRACK 5· JULY 2, 20260.95
    23. Graphs1.00

Transcript

229 cues· 3,160 words· 17,968 chars

  1. 0:12 Yes, my talk today is about the title, your data models, your moat.
  2. 0:16 We have an enterprise wide platform that we had just rolled out here this past month.
  3. 0:21 And so I'll go into details on this.
  4. 0:24 I'll give you some, hopefully some practical lessons here and why we made decisions we made for this, how you could picture your processes within a similar type of framework.
  5. 0:35 So first, just a quick introduction.
  6. 0:38 So this gets into the title here, the talk, and the framing of what I hope you take from this, but with AI moving very fast at the frontier, what's defensible?
  7. 0:49 You can move, you can build things very quickly with clogged code, but once you push things to production, there's constraints.
  8. 0:59 You find how much of your deployed stack do you want to actually own?
  9. 1:04 There's monitoring, there's upkeep,
  10. 1:06 You know, people building dependencies off your stack that you have to be prepared to handle.
  11. 1:11 How much appetite do users have for decentralized access?
  12. 1:14 This gets into, I'll show you what we built, but you know, what's the access point for users?
  13. 1:21 Is it another chat app?
  14. 1:22 Is it Claude?
  15. 1:24 Is it ChatGPT?
  16. 1:26 Is it something else?
  17. 1:28 What's your product differentiation from those different SaaS products?
  18. 1:32 And so our team then, with this context in mind, thought through here, what's our skill set here?
  19. 1:39 What's our competitive advantage in this environment?
  20. 1:42 And this is what I really hope that you take from this talk and picture yourself in this.
  21. 1:47 But our moat here was our understanding of our internal processes, the tacit knowledge that you need to run successful AI.
  22. 1:57 And this is true, I think, no matter how good
  23. 2:00 AI gets, how good models get in new releases that different companies put out when, when a mythos comes out or when there's a new app from Claudia, I'm not, I'm not worried because the part that we've built is the defensible part that that's, that's durable.
  24. 2:15 So these are the, I'll tell you what this means here in more detail, but these are the processes, tacit knowledge that we've modeled into what we call the strategic intelligence platform or SIP.
  25. 2:26 And it rolled out here this past month in production for enterprise use across the Gates Foundation, so about 4,000 people.
  26. 2:36 So first, I know this is an engineering talk, but the scope of this talk gets into data modeling, internal operations processes.
  27. 2:44 And so I wanna give very quick background here over what the Gates Foundation does, because then this is what we're modeling.
  28. 2:52 So as you're probably familiar, the Gates Foundation has a very wide scope and it's a very ambitious work that we've been doing for the past 25 plus years.
  29. 3:02 And there's all kinds of broad initiatives that we're doing, whether it's for child mortality, whether it's for nutrition, agriculture, education.
  30. 3:14 And these are kind of broadly the different buckets that these different initiatives fit into.
  31. 3:19 Creating market incentives, spurring innovation,
  32. 3:21 collaboration between public and private sectors.
  33. 3:24 And then the fourth one here kind of gets into the lens that we're building here.
  34. 3:30 High-quality data, trying to derive data-driven insights from the actual investments, the grants that we've put out.
  35. 3:39 And over 25 years, there's a ton of structure, there's a ton of data that's developed, and trying to extract those insights at scale is difficult, and that's what we're trying to solve.
  36. 3:52 So this slide here is a snapshot of some of the different – of the work that went out in 2023 within the foundation.
  37. 4:01 This gives you an idea that I just put this here to show some of the structure that we have that we're working across.
  38. 4:08 So you have over 2,000 grants in one year.
  39. 4:11 Many of these are 5 million plus.
  40. 4:14 Many hundred plus countries that are targeted with these grants.
  41. 4:20 alumni, so there's 4,000 different employees of the foundation, many different strategies within the foundation, the U.S., within the U.S., across almost all the states, grantees, the total annual disbursement over $7 billion.
  42. 4:37 And so this gives you some idea of structure that we're working with.
  43. 4:40 And this one just finally here, when I show the data model, this will make more sense, but we have different divisions that funding goes out through different divisions.
  44. 4:49 And so this breaks down some of those divisions that you can see different priorities and it'll make more sense in a second here, but global development, global health, gender equality, USP are just a sample of the different divisions.
  45. 5:05 Okay, so the fun stuff here now, I hope.
  46. 5:07 The strategic intelligence platform, so in a nutshell here, structuring operational data for agentic retrieval.
  47. 5:17 So we're building a knowledge graph with the idea of the agent consumer.
  48. 5:24 And here is an end-to-end look of what this looks like.
  49. 5:27 So we have different systems of record, structured, unstructured.
  50. 5:33 These have been siloed traditionally.

Chapters

  1. 0:00 The moat question: what stays defensible as AI commoditizes
  2. 2:32 SIP across the 4,000 person Gates Foundation
  3. 3:46 The scale: 25 years, $7B a year, 2,000 grants
  4. 5:15 Structuring operational data for agentic retrieval
  5. 6:18 Tacit knowledge as the moat: engaging data owners
  6. 7:09 The curation pipeline and governance
  7. 8:30 Modeling hierarchies: funding and management lenses
  8. 12:32 People, org charts, and stitching siloed systems
  9. 13:49 Combining unstructured documents with structured data
  10. 15:34 Serving the graph to Claude through MCP
  11. 17:31 Retrieval evals and the feedback loop

Open at this second