read-only demo

Videos DqtmZE6Hl0g

The Prompt is the Platform - Dominik Tornow, Resonate HQ

index_state ready data_status ok

AI Engineer· published 2026-06-29· 0:17:33· en-US· indexed 2026-08-11 05:48

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:14, 1 of 1 keyframes kept
  2. Shot 1, 0:14 to 0:31, 1 of 1 keyframes kept
  3. Shot 2, 0:31 to 0:56, 1 of 1 keyframes kept
  4. Shot 3, 0:56 to 1:26, 1 of 1 keyframes kept
  5. Shot 4, 1:26 to 1:56, 0 of 1 keyframes kept
  6. Shot 5, 1:56 to 2:26, 1 of 1 keyframes kept
  7. Shot 6, 2:26 to 2:56, 0 of 1 keyframes kept
  8. Shot 7, 2:56 to 3:21, 0 of 1 keyframes kept
  9. Shot 8, 3:21 to 3:40, 1 of 1 keyframes kept
  10. Shot 9, 3:40 to 4:06, 0 of 1 keyframes kept
  11. Shot 10, 4:06 to 4:32, 0 of 1 keyframes kept
  12. Shot 11, 4:32 to 4:58, 0 of 1 keyframes kept
  13. Shot 12, 4:58 to 5:24, 0 of 1 keyframes kept
  14. Shot 13, 5:24 to 5:50, 1 of 1 keyframes kept
  15. Shot 14, 5:50 to 6:26, 0 of 1 keyframes kept
  16. Shot 15, 6:26 to 7:02, 0 of 1 keyframes kept
  17. Shot 16, 7:02 to 7:30, 1 of 1 keyframes kept
  18. Shot 17, 7:30 to 7:58, 0 of 1 keyframes kept
  19. Shot 18, 7:58 to 8:27, 0 of 1 keyframes kept
  20. Shot 19, 8:27 to 8:54, 0 of 1 keyframes kept
  21. Shot 20, 8:54 to 9:21, 0 of 1 keyframes kept
  22. Shot 21, 9:21 to 9:53, 0 of 1 keyframes kept
  23. Shot 22, 9:53 to 10:22, 1 of 1 keyframes kept
  24. Shot 23, 10:22 to 10:50, 0 of 1 keyframes kept
  25. Shot 24, 10:50 to 11:18, 0 of 1 keyframes kept
  26. Shot 25, 11:18 to 11:44, 1 of 1 keyframes kept
  27. Shot 26, 11:44 to 12:09, 0 of 1 keyframes kept
  28. Shot 27, 12:09 to 12:35, 1 of 1 keyframes kept
  29. Shot 28, 12:35 to 13:00, 0 of 1 keyframes kept
  30. Shot 29, 13:00 to 13:25, 0 of 1 keyframes kept
  31. Shot 30, 13:25 to 13:51, 0 of 1 keyframes kept
  32. Shot 31, 13:51 to 14:27, 1 of 1 keyframes kept
  33. Shot 32, 14:27 to 15:03, 0 of 1 keyframes kept
  34. Shot 33, 15:03 to 15:39, 1 of 1 keyframes kept
  35. Shot 34, 15:39 to 16:14, 0 of 1 keyframes kept
  36. Shot 35, 16:14 to 16:59, 1 of 1 keyframes kept
  37. Shot 36, 16:59 to 17:32, 1 of 1 keyframes kept

37 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
190
whisperx 190
chunks
32
from 190 cues
keyframes
15
kept of 37 captured
frames with text
15
99 lines read
chapters
0
from the source metadata
keyframe bytes
1.8 MB
word timings on 190 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done 2026-08-11 05:45 1m 17s
stt done 2026-08-11 05:47 16s
chunk done 2026-08-11 05:47 0s
text_embed done 2026-08-11 05:47 0s
keyframe done 2026-08-11 05:47 49s
ocr done 2026-08-11 05:48 3s
frame_embed done 2026-08-11 05:48 3s

Frames, and what the machine read

  • 0:04 #0 done3 line(s)

    shot 0·sharpness 718.8

    1. The Prompt is the Platform1.00
    2. * Resonating...0.96
    3. Al Engineer0.98
  • 0:21 #1 done2 line(s)

    shot 1·sharpness 877.4

    1. Dominik Tornow1.00
    2. Resonate HQ1.00
  • 0:53 #2 done2 line(s)

    shot 2·sharpness 480.6

    1. general1.00
    2. bespoke1.00
  • 1:16 #3 done4 line(s)

    shot 3·sharpness 907.9

    1. implementation1.00
    2. specification1.00
    3. implementation1.00
    4. implementa0.99
  • 1:52 #4 skipped

    shot 4·duplicate of #3

  • 2:22 #5 done4 line(s)

    shot 5·sharpness 964.3

    1. Resonate Server1.00
    2. Resonate Server0.99
    3. Resonate Protocol1.00
    4. Resonate Se1.00
  • 2:32 #6 skipped

    shot 6·duplicate of #3

  • 3:18 #7 skipped

    shot 7·duplicate of #2

  • 3:34 #8 done2 line(s)

    shot 8·sharpness 923.0

    1. Resonate Durable Execution1.00
    2. on NATS.IO1.00
  • 3:58 #9 skipped

    shot 9·duplicate of #2

  • 4:29 #10 skipped

    shot 10·duplicate of #2

  • 4:55 #11 skipped

    shot 11·duplicate of #2

  • 5:16 #12 skipped

    shot 12·duplicate of #2

  • 5:40 #13 done6 line(s)

    shot 13·sharpness 854.6

    1. Target1.00
    2. Spec1.00
    3. Abstract1.00
    4. Concrete1.00
    5. Spec1.00
    6. Impl1.00
  • 6:01 #14 skipped

    shot 14·duplicate of #2

  • 6:47 #15 skipped

    shot 15·duplicate of #2

  • 7:21 #16 done11 line(s)

    shot 16·sharpness 1535.0

    1. Agent1.00
    2. Agent1.00
    3. Agent1.00
    4. Abstract1.00
    5. Simulation1.00
    6. Concrete1.00
    7. Concrete1.00
    8. Spec1.00
    9. Impl0.99
    10. Spec1.00
    11. Impl1.00
  • 7:52 #17 skipped

    shot 17·duplicate of #3

  • 8:13 #18 skipped

    shot 18·duplicate of #16

  • 8:45 #19 skipped

    shot 19·duplicate of #2

  • 9:00 #20 skipped

    shot 20·duplicate of #2

  • 9:37 #21 skipped

    shot 21·duplicate of #2

  • 9:57 #22 done12 line(s)

    shot 22·sharpness 1916.8

    1. >>> kvs = KVStore(...)0.98
    2. >>> kvs = KVStore(...)0.94
    3. >>> t = kvs.create("key", "foo")0.99
    4. >>> t = kvs.create("key", "foo")0.99
    5. >>> t = kvs.update("key", "bar", t)0.97
    6. >>> t = kvs.update("key", "bar", t)0.97
    7. >>> t = kvs.update("key", "baz", t)0.99
    8. >>> t = kvs.update("key", "baz", t)0.99
    9. >>> kvs.get("key")0.98
    10. >>> kvs.get("key")0.99
    11. (“baz", 2)#FRESH0.90
    12. ("foo", 0)#STALE0.93
  • 10:33 #23 skipped

    shot 23·duplicate of #22

Transcript

190 cues· 2,081 words· 12,329 chars

  1. 0:02 In 2026, coding agents will quietly retire their first software platform.
  2. 0:09 Not because it's bad, simply because the platform is unnecessary.
  3. 0:15 I am Dominik Thurnow.
  4. 0:16 I am founder and CEO of Resonate.
  5. 0:19 Resonate is a durable execution platform built with minimalism and simplicity as its core technical values.
  6. 0:27 And these properties will play a central role in this talk.
  7. 0:31 At Resonate, we have a working theory where software engineering is headed.
  8. 0:37 General purpose implementations will increasingly be replaced by bespoke implementations generated on demand, not as a new library, a new framework, or a new platform, but as a minimal extension of the infrastructure that is already in place.
  9. 0:57 If this theory holds true,
  10. 1:00 reuse will move upstream.
  11. 1:03 Instead of reusing a general purpose implementation, we will reuse a specification, and we will derive a bespoke implementation from it.
  12. 1:16 In fact, we can build many bespoke implementations, tailor-made for the infrastructure that is already in place.
  13. 1:24 We just have to ask the agent.
  14. 1:27 At this point, the prompt is the platform.
  15. 1:33 Resonate is a dual execution platform.
  16. 1:35 We have an implementation of the Resonate server.
  17. 1:38 We have implementations of the Resonate SDK for TypeScript, Python, Rust, Go, and Java.
  18. 1:45 So we have to ask, what does this new reality mean for us?
  19. 1:51 If implementations become generatable, where does our value live?
  20. 1:57 And our answer, our value moves from implementation to specification.
  21. 2:05 Now this changes how we think about Resonate.
  22. 2:09 The product is no longer the implementation.
  23. 2:12 The product is the specification, the protocol.
  24. 2:15 And from that protocol, we want to derive multiple server implementations.
  25. 2:22 One is a general purpose Resonate server, our reference implementation.
  26. 2:26 Others are implementations built with infrastructure partners.
  27. 2:31 For customers and partners, this means durable execution right on top of their existing infrastructure with minimal additional dependencies.
  28. 2:40 So the question is no longer, can we build a server?
  29. 2:44 The question is, can we repeatedly synthesize trusted servers from the same specification?
  30. 2:51 And if so, how?
  31. 2:58 When we talk about agentic engineering, we focus all of our attention on verification.
  32. 3:04 How do we know the result is correct?
  33. 3:07 But today, I want to focus on the specification instead.
  34. 3:12 And more importantly, how can agents participate in specifying the system, not just building or verifying it?
  35. 3:23 Now, Resonate is partnering with multiple infrastructure providers to bring durable executions natively to their technology stack.
  36. 3:31 One of them is Sunadia, the company behind NATSIO, an open source messaging system designed for building modern distributed systems.
  37. 3:41 For the rest of this presentation, we will use Resonate on NATSIO to explore our agentic engineering practices.
  38. 3:49 How do we go from specification to implementation?
  39. 3:53 First, we need to level set our mental model.
  40. 3:58 This picture is a common view of agent decoding.
  41. 4:02 There's an agent, there's a specification, and then there's an implementation.
  42. 4:08 And for many applications, that is enough.
  43. 4:11 But it is not enough for what we are trying to do, because we are not trying to generate one implementation from a specification.
  44. 4:23 We are trying to generate multiple target-specific implementations from the specification.
  45. 4:29 So the specification must not take any aspect of an implementation into account.
  46. 4:36 The specification must not assume a concrete database schema or concrete indices.
  47. 4:42 The specification must not even assume a relational database with tables and transactions at all.
  48. 4:48 It must not assume a key value store.
  49. 4:50 It must not assume weak consistency.
  50. 4:52 It must not assume strong consistency.

Open at this second