vidtheque.

read-only demo

Videos whue9_YquGA

Building an Autonomous Engineering Org - Angie Jones, Agentic AI Foundation

index_state ready data_status ok

AI Engineer· published 2026-06-28· 0:17:36· en-US· indexed 2026-08-11 08:01

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:26, 1 of 1 keyframes kept
  2. Shot 1, 0:26 to 0:53, 1 of 1 keyframes kept
  3. Shot 2, 0:53 to 1:19, 0 of 1 keyframes kept
  4. Shot 3, 1:19 to 1:46, 1 of 1 keyframes kept
  5. Shot 4, 1:46 to 2:12, 0 of 1 keyframes kept
  6. Shot 5, 2:12 to 2:39, 1 of 1 keyframes kept
  7. Shot 6, 2:39 to 3:05, 1 of 1 keyframes kept
  8. Shot 7, 3:05 to 3:32, 1 of 1 keyframes kept
  9. Shot 8, 3:32 to 3:59, 0 of 1 keyframes kept
  10. Shot 9, 3:59 to 4:25, 1 of 1 keyframes kept
  11. Shot 10, 4:25 to 4:52, 1 of 1 keyframes kept
  12. Shot 11, 4:52 to 5:19, 1 of 1 keyframes kept
  13. Shot 12, 5:19 to 5:46, 0 of 1 keyframes kept
  14. Shot 13, 5:46 to 6:13, 1 of 1 keyframes kept
  15. Shot 14, 6:13 to 6:41, 1 of 1 keyframes kept
  16. Shot 15, 6:41 to 7:08, 0 of 1 keyframes kept
  17. Shot 16, 7:08 to 7:34, 1 of 1 keyframes kept
  18. Shot 17, 7:34 to 8:01, 0 of 1 keyframes kept
  19. Shot 18, 8:01 to 8:27, 1 of 1 keyframes kept
  20. Shot 19, 8:27 to 8:54, 0 of 1 keyframes kept
  21. Shot 20, 8:54 to 9:20, 0 of 1 keyframes kept
  22. Shot 21, 9:20 to 9:47, 1 of 1 keyframes kept
  23. Shot 22, 9:47 to 10:13, 1 of 1 keyframes kept
  24. Shot 23, 10:13 to 10:40, 0 of 1 keyframes kept
  25. Shot 24, 10:40 to 11:06, 1 of 1 keyframes kept
  26. Shot 25, 11:06 to 11:33, 0 of 1 keyframes kept
  27. Shot 26, 11:33 to 11:59, 1 of 1 keyframes kept
  28. Shot 27, 11:59 to 12:26, 1 of 1 keyframes kept
  29. Shot 28, 12:26 to 12:53, 0 of 1 keyframes kept
  30. Shot 29, 12:53 to 13:06, 1 of 1 keyframes kept
  31. Shot 30, 13:06 to 13:31, 1 of 1 keyframes kept
  32. Shot 31, 13:31 to 13:56, 1 of 1 keyframes kept
  33. Shot 32, 13:56 to 14:21, 0 of 1 keyframes kept
  34. Shot 33, 14:21 to 14:46, 1 of 1 keyframes kept
  35. Shot 34, 14:46 to 15:14, 1 of 1 keyframes kept
  36. Shot 35, 15:14 to 15:43, 1 of 1 keyframes kept
  37. Shot 36, 15:43 to 16:11, 0 of 1 keyframes kept
  38. Shot 37, 16:11 to 16:41, 1 of 1 keyframes kept
  39. Shot 38, 16:41 to 17:07, 1 of 1 keyframes kept
  40. Shot 39, 17:07 to 17:09, 1 of 1 keyframes kept
  41. Shot 40, 17:09 to 17:11, 1 of 1 keyframes kept
  42. Shot 41, 17:11 to 17:14, 0 of 1 keyframes kept
  43. Shot 42, 17:14 to 17:15, 1 of 1 keyframes kept
  44. Shot 43, 17:15 to 17:19, 1 of 1 keyframes kept
  45. Shot 44, 17:19 to 17:22, 0 of 1 keyframes kept
  46. Shot 45, 17:22 to 17:25, 1 of 1 keyframes kept
  47. Shot 46, 17:25 to 17:28, 1 of 1 keyframes kept
  48. Shot 47, 17:28 to 17:30, 0 of 1 keyframes kept
  49. Shot 48, 17:30 to 17:35, 1 of 1 keyframes kept

49 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
168
whisperx 168
chunks
31
from 168 cues
keyframes
33
kept of 49 captured
frames with text
26
200 lines read
chapters
0
from the source metadata
keyframe bytes
6.6 MB
word timings on 168 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-11 07:57 1m 17s
stt done — 2026-08-11 07:58 18s
chunk done — 2026-08-11 07:59 0s
text_embed done — 2026-08-11 07:59 1s
keyframe done — 2026-08-11 07:59 1m 44s
ocr done — 2026-08-11 08:00 7s
frame_embed done — 2026-08-11 08:00 7s

Frames, and what the machine read

  • 0:10 #0 done5 line(s)

    shot 0·sharpness 724.8

    1. ANGIE JONES, VP OF AGENTIC AI FOUNDATION0.97
    2. Building an1.00
    3. Autonomous1.00
    4. Engineering Org1.00
    5. angiejones.tech1.00

  • 0:47 #1 done4 line(s)

    shot 1·sharpness 489.8

    1. THE TURNING POINT1.00
    2. Code is not the1.00
    3. bottleneck.1.00
    4. angiejones.tech1.00

  • 1:06 #2 skipped

    shot 2·duplicate of #1

  • 1:25 #3 done12 line(s)

    shot 3·sharpness 696.6

    1. Where We Are1.00
    2. Experimentation1.00
    3. Adoption1.00
    4. Impact1.00
    5. explore and normalize Al0.99
    6. scale usage through1.00
    7. drive measurable outcomes1.00
    8. usage1.00
    9. structure and consistency1.00
    10. and productivity gains1.00
    11. ARE0.78
    12. HERE1.00

  • 1:49 #4 skipped

    shot 4·duplicate of #3

  • 2:33 #5 done4 line(s)

    shot 5·sharpness 459.5

    1. AGENTIC ENGINEER1.00
    2. We need to be1.00
    3. Agentic Engineers.1.00
    4. angiejones.tech1.00

  • 2:57 #6 done4 line(s)

    shot 6·sharpness 461.9

    1. AGENTIC ENGINEER1.00
    2. Treat agents as0.98
    3. core collaborators.1.00
    4. angiejones.tech1.00

  • 3:21 #7 done10 line(s)

    shot 7·sharpness 766.9

    1. AGENTIC ENGINEER1.00
    2. Al Maturity Model0.97
    3. Stage 00.97
    4. Stage 10.99
    5. Stage 21.00
    6. Stage 31.00
    7. Stage 41.00
    8. Stage 51.00
    9. adapted from Steve Yegge's Gas Town article0.98
    10. angiejones.tech1.00

  • 3:55 #8 skipped

    shot 8·duplicate of #7

  • 4:14 #9 done16 line(s)

    shot 9·sharpness 1089.2

    1. AGENTIC ENGINEER1.00
    2. Al Maturity Model0.98
    3. Stage 00.96
    4. Stage 10.99
    5. Stage 21.00
    6. Stage 31.00
    7. Stage 41.00
    8. Stage 51.00
    9. Unengaged1.00
    10. Assisted1.00
    11. Conversational1.00
    12. Directed1.00
    13. Parallel1.00
    14. Autonomous1.00
    15. adapted from Steve Yegge's Gas Town article0.99
    16. angiejones.tech1.00

  • 4:38 #10 done4 line(s)

    shot 10·sharpness 356.5

    1. AGENTIC ENGINEER1.00
    2. How can we1.00
    3. level up?0.98
    4. angiejones.tech1.00

  • 4:55 #11 done9 line(s)

    shot 11·sharpness 606.1

    1. AGENTIC ENGINEER1.00
    2. 1%0.98
    3. Power Users1.00
    4. 1/9/90 Rule1.00
    5. 9%0.96
    6. Tinkerers1.00
    7. Consumers1.00
    8. 90%1.00
    9. angiejones.tech1.00

  • 5:43 #12 skipped

    shot 12·duplicate of #11

  • 6:08 #13 done5 line(s)

    shot 13·sharpness 372.6

    1. AI CHAMPIONS0.98
    2. The 1%:1.00
    3. Al Champions0.97
    4. actually ~1.43%0.97
    5. angiejones.tech1.00

  • 6:35 #14 done15 line(s)

    shot 14·sharpness 913.5

    1. STAGE 1/21.00
    2. Al Maturity Model0.99
    3. Stage 01.00
    4. Stage 10.98
    5. Stage 21.00
    6. Stage 31.00
    7. Stage 41.00
    8. Stage 51.00
    9. Unengaged1.00
    10. Assisted1.00
    11. Conversational1.00
    12. Directed1.00
    13. Parallel1.00
    14. Autonomous1.00
    15. angiejones.tech1.00

  • 7:05 #15 skipped

    shot 15·duplicate of #14

  • 7:14 #16 done4 line(s)

    shot 16·sharpness 554.4

    1. STAGE 30.99
    2. Turning repos into1.00
    3. Al-friendly systems.1.00
    4. angiejones.tech1.00

  • 7:40 #17 skipped

    shot 17·duplicate of #16

  • 8:14 #18 done5 line(s)

    shot 18·sharpness 507.9

    1. STAGE 30.98
    2. Not all0.99
    3. codebases are1.00
    4. equal.1.00
    5. angiejones.tech1.00

  • 8:41 #19 skipped

    shot 19·duplicate of #18

  • 8:57 #20 skipped

    shot 20·duplicate of #18

  • 9:36 #21 done13 line(s)

    shot 21·sharpness 733.0

    1. STAGE 30.99
    2. Al Friendly Repo0.99
    3. Agent Context1.00
    4. Agent Rules1.00
    5. Al Workflows0.99
    6. Al Code Review1.00
    7. Al PR Indicators0.97
    8. 251.00
    9. 251.00
    10. 251.00
    11. 251.00
    12. 151.00
    13. angiejones.tech1.00

  • 10:00 #22 done15 line(s)

    shot 22·sharpness 727.4

    1. STAGE 31.00
    2. Al Maturity Model0.99
    3. Stage 00.99
    4. Stage 10.97
    5. Stage 21.00
    6. Stage 30.99
    7. Stage 41.00
    8. Stage 51.00
    9. Unengaged1.00
    10. Assisted1.00
    11. Conversational1.00
    12. Directed1.00
    13. Parallel1.00
    14. Autonomous1.00
    15. angiejones.tech1.00

  • 10:16 #23 skipped

    shot 23·duplicate of #22

Transcript

168 cues· 2,718 words· 15,149 chars

  1. 0:00 So I've spent the last couple of years transforming Block's 3,500-person engineering org into an autonomous one.
  2. 0:08 And this is a challenge that most tech companies are actively trying to solve or will be in the very near future.
  3. 0:14 So today, I'll share our path to getting there.
  4. 0:18 Our agentic coding journey started very early.
  5. 0:20 We were building Goose, our internal coding agent, even before the LLM supported tool calling.
  6. 0:27 We worked with Anthropic as design partners for the initial release of MCP, and Goose became the reference implementation for the MCP client.
  7. 0:35 So internally, our most curious engineers were some of the very first in the industry to use these types of coding agents.
  8. 0:45 After a couple of months, about 90% of our engineers were now using tools like Goose and Cloud Code regularly to generate code.
  9. 0:53 So on paper, it looked like we were all in on AI.
  10. 0:57 But our CEO was convinced that engineering wasn't using AI at all, like they couldn't be, right?
  11. 1:02 As far as he was concerned, because we weren't shipping any faster.
  12. 1:07 Well, I had the numbers, both the metrics and the token bills.
  13. 1:10 So I knew that engineering was in fact using AI, but he was right.
  14. 1:15 Features certainly weren't making it to our customers any faster.
  15. 1:18 So I began to dig into this a bit.
  16. 1:22 I think of AI enablements in three phases, experimentation, adoption, and impact.
  17. 1:28 I'd say that we surpassed the experimentation phase as 90% of our engineering org was using AI, but they still were kind of using it inside of the IDE, you know, asking questions or writing boilerplate code.
  18. 1:43 And I knew that if we really wanted to see impactful outcomes, then we needed a way to integrate AI into how we build and ship.
  19. 1:52 I'd spent the first half of 2025 leading AI enablement for our entire company.
  20. 1:58 So 12,000 employees across every function, marketing, design, finance, legal.
  21. 2:04 Now our CTO tasked me with building an energetic engineering org.
  22. 2:10 Okay, sure, but what does that actually mean?
  23. 2:13 There's no playbook for any of this stuff, right?
  24. 2:15 And I know because I went looking at your blogs, hoping that you all had it all figured out, but I only saw a bunch of posts saying how you all were making it up as you went along, so...
  25. 2:27 I did the same, right?
  26. 2:29 In the simplest of terms, I defined an agentic engineering org as one where engineers leverage AI agents as their primary means of producing engineering outcomes.
  27. 2:40 So that meant that our engineers needed to treat agents as core members of their engineering workflow, not just using AI to help them write code here and there, but actually working with agents, decomposing problems, delegating work, reviewing and verifying what was done.
  28. 2:59 And we wanted them directing the work as their default way of operating.
  29. 3:03 But of course, the vast majority of our engineers were not here.
  30. 3:08 So to see where we needed to go, I came up with the maturity model.
  31. 3:12 This model measures the engineer's relationship with AI agents.
  32. 3:16 So how they think and delegate and orchestrate.
  33. 3:20 And while I had some form of this model in Q3 of last year, Steve Yeagey's Gastown article helped me reorganize it into a better model.
  34. 3:30 So stage zero is where the engineer doesn't use AI tools in their workflow at all.
  35. 3:35 Stage one, they use AI to auto-complete, but they've never used it in agent mode maybe.
  36. 3:42 Stage two is where engineers are chatting with agents, but not using it to produce any PRs.
  37. 3:49 Stage three, engineers are delegating tasks to agents and reviewing the output.
  38. 3:54 Those at stage four are running multiple agents in parallel, and then stage five is that final boss where engineers are delegating complete tasks to agents, and the agent is able to produce shippable results without the human necessarily needing to guide it.
  39. 4:09 So I'd say by the end of the first half, the bulk of our engineers were between stages one and two, and I needed to get them to stage five.
  40. 4:18 Now, I wasn't quite sure how to get 3,500 engineers to that level, especially when one, this is all highly experimental, right?
  41. 4:27 There's no playbook, again, for any of this.
  42. 4:31 Two, things were moving so fast that what might be a best practice this week could be outdated when a new tool or a new model drops next week.
  43. 4:40 And honestly, this was leading to AI fatigue.
  44. 4:43 And then three, people were already feeling turned off by the top down pressure from leadership to essentially AI or die.
  45. 4:52 So I thought about the one nine ninety rule where in digital communities, about one percent create, nine percent interact and 90 percent passively consume.
  46. 5:02 And this maps almost perfectly to how engineers adopt AI.
  47. 5:06 So you'll have a small percentage that'll deep dive and start creating the genetic patterns and discovering useful techniques for working with agents.
  48. 5:15 You'll have some that might tweak the agent's MD file here or there, but most people aren't gonna spend the extra cycles to figure this stuff out.
  49. 5:25 So I realized that if my AI strategy depends on every individual leveling themselves up, I'm never going to see that broad impact.
  50. 5:35 So I leaned into this model and I used it to my advantage.

Open at this second