vidtheque.

read-only demo

Videos XNtkiQJ49Ps

Agents need more than a chat - Jacob Lauritzen, CTO Legora

index_state ready data_status ok

AI Engineer· published 2026-04-22· 0:14:21· en-US· indexed 2026-08-10 19:53

Open on YouTube

Scene timeline

  1. Shot 0, 0:00 to 0:05, 1 of 1 keyframes kept
  2. Shot 1, 0:05 to 0:09, 1 of 1 keyframes kept
  3. Shot 2, 0:09 to 0:15, 1 of 1 keyframes kept
  4. Shot 3, 0:15 to 0:17, 1 of 1 keyframes kept
  5. Shot 4, 0:17 to 0:37, 1 of 1 keyframes kept
  6. Shot 5, 0:37 to 0:50, 1 of 1 keyframes kept
  7. Shot 6, 0:50 to 0:56, 1 of 1 keyframes kept
  8. Shot 7, 0:56 to 0:59, 1 of 1 keyframes kept
  9. Shot 8, 0:59 to 1:02, 1 of 1 keyframes kept
  10. Shot 9, 1:02 to 1:04, 1 of 1 keyframes kept
  11. Shot 10, 1:04 to 1:07, 1 of 1 keyframes kept
  12. Shot 11, 1:07 to 1:14, 1 of 1 keyframes kept
  13. Shot 12, 1:14 to 1:25, 1 of 1 keyframes kept
  14. Shot 13, 1:25 to 1:28, 1 of 1 keyframes kept
  15. Shot 14, 1:28 to 1:30, 1 of 1 keyframes kept
  16. Shot 15, 1:30 to 1:37, 1 of 1 keyframes kept
  17. Shot 16, 1:37 to 1:41, 1 of 1 keyframes kept
  18. Shot 17, 1:41 to 1:53, 1 of 1 keyframes kept
  19. Shot 18, 1:53 to 2:09, 1 of 1 keyframes kept
  20. Shot 19, 2:09 to 2:10, 1 of 1 keyframes kept
  21. Shot 20, 2:10 to 2:23, 1 of 1 keyframes kept
  22. Shot 21, 2:23 to 2:27, 1 of 1 keyframes kept
  23. Shot 22, 2:27 to 2:38, 1 of 1 keyframes kept
  24. Shot 23, 2:38 to 2:50, 1 of 1 keyframes kept
  25. Shot 24, 2:50 to 2:59, 1 of 1 keyframes kept
  26. Shot 25, 2:59 to 3:08, 1 of 1 keyframes kept
  27. Shot 26, 3:08 to 3:09, 1 of 1 keyframes kept
  28. Shot 27, 3:09 to 3:14, 1 of 1 keyframes kept
  29. Shot 28, 3:14 to 3:24, 1 of 1 keyframes kept
  30. Shot 29, 3:24 to 3:25, 1 of 1 keyframes kept
  31. Shot 30, 3:25 to 3:42, 1 of 1 keyframes kept
  32. Shot 31, 3:42 to 4:01, 1 of 1 keyframes kept
  33. Shot 32, 4:01 to 4:03, 1 of 1 keyframes kept
  34. Shot 33, 4:03 to 4:17, 1 of 1 keyframes kept
  35. Shot 34, 4:17 to 4:37, 1 of 1 keyframes kept
  36. Shot 35, 4:37 to 4:48, 1 of 1 keyframes kept
  37. Shot 36, 4:48 to 5:04, 1 of 1 keyframes kept
  38. Shot 37, 5:04 to 5:44, 1 of 1 keyframes kept
  39. Shot 38, 5:44 to 5:50, 1 of 1 keyframes kept
  40. Shot 39, 5:50 to 5:53, 1 of 1 keyframes kept
  41. Shot 40, 5:53 to 6:08, 1 of 1 keyframes kept
  42. Shot 41, 6:08 to 6:32, 1 of 1 keyframes kept
  43. Shot 42, 6:32 to 7:08, 1 of 1 keyframes kept
  44. Shot 43, 7:08 to 7:21, 1 of 1 keyframes kept
  45. Shot 44, 7:21 to 7:34, 1 of 1 keyframes kept
  46. Shot 45, 7:34 to 7:52, 1 of 1 keyframes kept
  47. Shot 46, 7:52 to 7:55, 1 of 1 keyframes kept
  48. Shot 47, 7:55 to 8:38, 1 of 1 keyframes kept
  49. Shot 48, 8:38 to 8:39, 1 of 1 keyframes kept
  50. Shot 49, 8:39 to 9:08, 1 of 1 keyframes kept
  51. Shot 50, 9:08 to 9:36, 1 of 1 keyframes kept
  52. Shot 51, 9:36 to 9:48, 1 of 1 keyframes kept
  53. Shot 52, 9:48 to 10:12, 1 of 1 keyframes kept
  54. Shot 53, 10:12 to 10:53, 1 of 1 keyframes kept
  55. Shot 54, 10:53 to 11:19, 1 of 1 keyframes kept
  56. Shot 55, 11:19 to 11:24, 1 of 1 keyframes kept
  57. Shot 56, 11:24 to 11:29, 1 of 1 keyframes kept
  58. Shot 57, 11:29 to 11:35, 1 of 1 keyframes kept
  59. Shot 58, 11:35 to 11:41, 1 of 1 keyframes kept
  60. Shot 59, 11:41 to 11:58, 1 of 1 keyframes kept
  61. Shot 60, 11:58 to 12:14, 1 of 1 keyframes kept
  62. Shot 61, 12:14 to 12:18, 1 of 1 keyframes kept
  63. Shot 62, 12:18 to 12:21, 1 of 1 keyframes kept
  64. Shot 63, 12:21 to 12:32, 1 of 1 keyframes kept
  65. Shot 64, 12:32 to 12:44, 1 of 1 keyframes kept
  66. Shot 65, 12:44 to 12:55, 1 of 1 keyframes kept
  67. Shot 66, 12:55 to 13:01, 1 of 1 keyframes kept
  68. Shot 67, 13:01 to 13:34, 1 of 1 keyframes kept
  69. Shot 68, 13:34 to 13:41, 1 of 1 keyframes kept
  70. Shot 69, 13:41 to 14:06, 1 of 1 keyframes kept
  71. Shot 70, 14:06 to 14:20, 1 of 1 keyframes kept
  72. Shot 71, 14:20 to 14:20, 0 of 1 keyframes kept

72 shot(s).

keyframes kept every frame deduplicated

What was stored

cues
211
whisperx 211
chunks
26
from 211 cues
keyframes
71
kept of 72 captured
frames with text
71
2,057 lines read
chapters
0
from the source metadata
keyframe bytes
6.8 MB
word timings on 211 cues

Provenance

Each pipeline stage, its state and the model that produced it
stage state model started took
fetch done — 2026-08-10 17:35 2m 06s
stt done — 2026-08-10 17:37 15s
chunk done — 2026-08-10 17:37 0s
text_embed done — 2026-08-10 19:53 1s
keyframe done — 2026-08-10 17:37 1m 33s
ocr done — 2026-08-10 17:39 35s
frame_embed done — 2026-08-10 19:53 12s

Frames, and what the machine read

  • 0:03 #0 done2 line(s)

    shot 0·sharpness 666.8

    1. AlEngineer0.98
    2. EUROPE1.00

  • 0:07 #1 done2 line(s)

    shot 1·sharpness 825.0

    1. PRESENTING SPONSOR0.98
    2. Google DeepMind1.00

  • 0:13 #2 done3 line(s)

    shot 2·sharpness 909.1

    1. PLATINUM SPONSORS0.98
    2. # Braintrust0.95
    3. WorkOS OpenAI0.95

  • 0:17 #3 done6 line(s)

    shot 3·sharpness 168.1

    1. Agents need more0.98
    2. than chat1.00
    3. LE0.95
    4. LEGORA1.00
    5. AlEngineer0.96
    6. d0.99

  • 0:31 #4 done14 line(s)

    shot 4·sharpness 1591.4

    1. gineer1.00
    2. ROPE0.98
    3. Jst0.87
    4. Agents need more1.00
    5. than chat1.00
    6. r.del0.77
    7. LEGORA1.00
    8. AlEngineer0.97
    9. AGENTS NEED MORE THAN A CHAT1.00
    10. EUROPE1.00
    11. PRESENTED BY1.00
    12. Google DeepMind1.00
    13. JACOB LAURITZEN/ CTO0.98
    14. LEGORA1.00

  • 0:38 #5 done19 line(s)

    shot 5·sharpness 1260.6

    1. LEGORA1.00
    2. COLLABORATIVE AI FOR LAWYERS1.00
    3. Research relevant law and draft a0.99
    4. contract between the two parties1.00
    5. in these documents.1.00
    6. Make no mistakes0.99
    7. AVEngi0.88
    8. Go0.68
    9. EUROI0.97
    10. Mic1.00
    11. AIEngine0.90
    12. EUROPE1.00
    13. AlEngineer0.99
    14. AGENTS NEED MORE THAN A CHAT1.00
    15. EUROPE1.00
    16. PRESENTED BY1.00
    17. Google DeepMind1.00
    18. JACOB LAURITZEN/CTO0.99
    19. LEGORA1.00

  • 0:51 #6 done29 line(s)

    shot 6·sharpness 1672.8

    1. LEGORA1.00
    2. COLLABORATIVE AI FOR LAWYERS0.99
    3. AlEngin0.94
    4. Research relevant law and draft a1.00
    5. EUROPE1.00
    6. contract between the two parties1.00
    7. in these documents.1.00
    8. Make no mistakes1.00
    9. Thinking...0.99
    10. en0.93
    11. Reading COMPANY_A.txt...1.00
    12. Reading COMPANY_B.txt...0.99
    13. Launching legal research sub agent...0.99
    14. ngin1.00
    15. Searching the web for "term 1"...0.97
    16. EUROPE1.00
    17. Searching in Swedish jurisdiction for "term 1"...0.99
    18. Searching in British jurisdiction for "term 1"...1.00
    19. Synthesizing...1.00
    20. Te1.00
    21. Thinking...1.00
    22. Searching the web for "term 2".0.99
    23. AlEngineer0.97
    24. AGENTS NEED MORE THAN A CHAT1.00
    25. EUROPE1.00
    26. PRESENTED BY1.00
    27. Google DeepMind1.00
    28. JACOB LAURITZEN/ CTO0.98
    29. LEGORA1.00

  • 0:58 #7 done29 line(s)

    shot 7·sharpness 1720.2

    1. Searching the web for "term 2"...0.98
    2. LEGORA1.00
    3. Searching British law for "term 2"...1.00
    4. eer0.92
    5. Thinking...0.99
    6. Writing LEGAL_RESEARCH_ANALYSIS.txt...1.00
    7. Legal research finished.1.00
    8. Researching company A...0.99
    9. Searching the web for "company A shareholders"...0.98
    10. Searching the web for "company A terms"...0.98
    11. Searching files for "company A"...0.99
    12. eer1.00
    13. Mi1.00
    14. Found 20 files. Reading...1.00
    15. Reading file "company a_shareholders.txt"...0.99
    16. Reading file "company a_terms.txt"...0.98
    17. r.de1.00
    18. AIEn0.98
    19. Reading folder "company a security0.99
    20. whitepapers"...0.99
    21. EU0.67
    22. Writing COMPANY A RESEARCH.txt..0.97
    23. AlEngineer0.99
    24. AGENTS NEED MORE THAN A CHAT1.00
    25. EUROPE1.00
    26. PRESENTED BY1.00
    27. Google DeepMind1.00
    28. JACOB LAURITZEN/ CTO0.98
    29. LEGORA1.00

  • 1:01 #8 done34 line(s)

    shot 8·sharpness 1887.5

    1. LEGORA1.00
    2. Company A research finished.1.00
    3. tog1.00
    4. Researching company B...0.99
    5. Searching the web for "company B shareholders"..0.98
    6. AEngineer0.97
    7. AlEng0.97
    8. EUROPE1.00
    9. Searching the web for "company B terms"...1.00
    10. Searching files for "company B"...0.99
    11. neo4i0.93
    12. SE0.69
    13. Found 20 files. Reading...0.99
    14. Engineer1.00
    15. Reading file "company a_shareholders.txt"..0.98
    16. EUROPE1.00
    17. AlEng0.96
    18. ESENTED BY1.00
    19. Reading file "company a_terms.txt"...0.98
    20. gle DeepMind1.00
    21. BeN0.77
    22. Reading folder "company a security0.97
    23. whitepapers"...1.00
    24. Writing COMPANY_B_RESEARCH.txt...1.00
    25. Company B research finished.1.00
    26. Looking up previous similar contracts...1.00
    27. Searching for similar previous engagements...0.99
    28. AlEngineer0.98
    29. AGENTS NEED MORE THAN A CHAT1.00
    30. EUROPE1.00
    31. PRESENTED BY1.00
    32. Google DeepMind1.00
    33. JACOB LAURITZEN/ CTO0.98
    34. LEGORA1.00

  • 1:02 #9 done27 line(s)

    shot 9·sharpness 1446.5

    1. LEGORA1.00
    2. Found 3 similar previous engagements.0.99
    3. AlEngin0.87
    4. together0.99
    5. Searching for non-solicitation clauses...1.00
    6. Found 2 clauses.0.99
    7. Op1.00
    8. AlEngineer0.96
    9. Previous contract lookup finished.1.00
    10. Thinking...0.96
    11. o4i0.97
    12. AIE1.00
    13. SENTR0.99
    14. neer1.00
    15. AlEngineer0.99
    16. ED BY1.00
    17. eepMind0.99
    18. eNCOR0.92
    19. AlEngine0.99
    20. EUROFE0.79
    21. AlEngineer0.98
    22. AGENTS NEED MORE THAN A CHAT1.00
    23. EUROPE1.00
    24. PRESENTED BY1.00
    25. Google DeepMind1.00
    26. JACOB LAURITZEN/ CTO0.98
    27. LEGORA1.00

  • 1:04 #10 done35 line(s)

    shot 10·sharpness 2410.7

    1. LEGORA1.00
    2. Found 3 similar previous engagements.0.99
    3. togethe0.96
    4. Searching for non-solicitation clauses...0.99
    5. Found 2 clauses.0.99
    6. enAl0.90
    7. AIE0.99
    8. Snorkel0.99
    9. Previous contract lookup finished.1.00
    10. Thinking...1.00
    11. AGREEMENT BETWEEN COMPANY A AND COMPANY B0.99
    12. Microsoft0.99
    13. Tessl0.95
    14. AlEng1.00
    15. This Agreement ("Agreement") is made and entered into as of [Effective Date], by and0.99
    16. between:1.00
    17. BeN0.63
    18. NIEng0.85
    19. place of business at [Address] ("Company A"); and0.98
    20. Company A, a company organized under the laws of [Jurisdiction], with its principal0.99
    21. Company B, a company organized under the laws of [Jurisdiction], with its principal0.99
    22. place of business at [Address] ("Company B").1.00
    23. Company A and Company B may be referred to individually as a "Party" and0.98
    24. collectively as the "Parties."1.00
    25. 1. PURPOSE0.94
    26. The purpose of this Agreement is to set forth the terms and conditions under which0.99
    27. the Parties will [describe purpose, e.g., collaborate on a project, provide services, or0.98
    28. supply goods].0.97
    29. AlEngineer0.97
    30. AGENTS NEED MORE THAN A CHAT1.00
    31. EUROPE1.00
    32. PRESENTED BY1.00
    33. Google DeepMind1.00
    34. JACOB LAURITZEN/CTO1.00
    35. LEGORA1.00

  • 1:08 #11 done29 line(s)

    shot 11·sharpness 1793.7

    1. LEGORA1.00
    2. 12.FORCE MAJEURE0.97
    3. Neither Party shall be liable for failure to perform due to events beyond its reasonable0.99
    4. togetherai0.99
    5. control.1.00
    6. 13.ENTIRE AGREEMENT0.99
    7. This Agreement constitutes the entire agreement between the Parties and0.98
    8. AlEngineer0.99
    9. rkel0.99
    10. supersedes all prior agreements.0.99
    11. EUROPE0.89
    12. 14.AMENDMENTS1.00
    13. SENTRY1.00
    14. Any amendment must be in writing and signed by both Parties.0.99
    15. AlEngineer1.00
    16. oft1.00
    17. EUROPE0.95
    18. Clause 3 doesn't look right.0.98
    19. Please use the precedent in the1.00
    20. eNCOR0.86
    21. previous "Company C" and0.98
    22. "Company D" contract0.97
    23. AlEngineer0.98
    24. AGENTS NEED MORE THAN A CHAT1.00
    25. EUROPE1.00
    26. PRESENTED BY1.00
    27. Google DeepMind1.00
    28. JACOB LAURITZEN/CTO1.00
    29. LEGORA1.00

  • 1:19 #12 done4 line(s)

    shot 12·sharpness 153.3

    1. LEGORA1.00
    2. COLLABORATIVE AI FOR LAWYERS0.98
    3. You're absolutely right!1.00
    4. Compacting history...1.00

  • 1:28 #13 done11 line(s)

    shot 13·sharpness 698.2

    1. LEGORA1.00
    2. COLLABORATIVE AI FOR LAWYERS1.00
    3. You're absolutely right!1.00
    4. Compacting history...1.00
    5. Researching company C...1.00
    6. Searching the web for “company C shareholders"...0.99
    7. Searching the web for "company C terms"...0.98
    8. Searching files for "company C"...0.98
    9. Found 5 files. Reading...1.00
    10. Reading file "company c_shareholders.txt"...0.97
    11. Reading file “company a terms.txt"...0.97

  • 1:30 #14 done20 line(s)

    shot 14·sharpness 2685.9

    1. Searching for similar previous engagements...0.99
    2. LEGORA1.00
    3. Found 3 similar previous engagements.1.00
    4. Searching for non-solicitation clauses...0.99
    5. Found 2 clauses.1.00
    6. Previous contract lookup finished.1.00
    7. Thinking...0.98
    8. AGREEMENT BETWEEN COMPANY A AND COMPANY B1.00
    9. This Agreement ("Agreement") is made and entered into as of [Effective Date], by and0.99
    10. between:1.00
    11. Company A, a company organized under the laws of [Jurisdiction], with its principal0.99
    12. place of business at [Address] ("Company A"); and0.97
    13. Company B, a company organized under the laws of [Jurisdiction], with its principal0.99
    14. place of business at [Address] ("Company B").1.00
    15. Company A and Company B may be referred to individually as a "Party" and1.00
    16. collectively as the "Parties."1.00
    17. 1.PURPOSE1.00
    18. The purpose of this Agreement is to set forth the terms and conditions under which0.99
    19. the Parties will [describe purpose, e.g., collaborate on a project, provide services, or1.00
    20. supply goodsl.0.98

  • 1:36 #15 done10 line(s)

    shot 15·sharpness 1126.1

    1. LEGORA1.00
    2. 12.FORCE MAJEURE0.99
    3. Neither Party shall be liable for failure to perform due to events beyond its reasonable0.99
    4. control.1.00
    5. 13.ENTIRE AGREEMENT0.98
    6. This Agreement constitutes the entire agreement between the Parties and0.99
    7. supersedes all prior agreements.0.99
    8. 14.AMENDMENTS1.00
    9. Any amendment must be in writing and signed by both Parties.1.00
    10. Did you also change clause 4???0.98

  • 1:40 #16 done15 line(s)

    shot 16·sharpness 1399.5

    1. PE0.97
    2. Eng1.00
    3. Not the greatest experience.0.99
    4. EURO0.84
    5. er1.00
    6. O0.91
    7. Aleng0.94
    8. EUR1.00
    9. AlEngineer0.97
    10. AGENTS NEED MORE THAN A CHAT0.98
    11. EUROPE1.00
    12. PRESENTED BY1.00
    13. Google DeepMind1.00
    14. JACOB LAURITZEN/ CTO0.98
    15. LEGORA1.00

  • 1:51 #17 done22 line(s)

    shot 17·sharpness 2403.2

    1. EU1.00
    2. LEGORA1.00
    3. COLLABORATIVE AI FOR LAWYERS0.99
    4. Legora is a collaborative Al workspace1.00
    5. for legal professionals0.98
    6. ineer0.99
    7. 1000+1.00
    8. $5.5bn+1.00
    9. $100m+1.00
    10. rose0.87
    11. ineer0.99
    12. Customers1.00
    13. Valuation1.00
    14. ARR1.00
    15. PE0.99
    16. AlEngineer0.98
    17. AGENTS NEED MORE THAN A CHAT1.00
    18. EUROPE1.00
    19. PRESENTED BY1.00
    20. Google DeepMind1.00
    21. JACOB LAURITZEN/ CTO0.98
    22. LEGORA1.00

  • 2:02 #18 done43 line(s)

    shot 18·sharpness 2446.2

    1. LEGORA1.00
    2. COLLABORATIVE AIFOR LAWYERS1.00
    3. Years from 1 to 10oM ARR: Direct Sales-Led GTM Motion0.97
    4. AlEng0.92
    5. Public & Private Workflow SaaS and AI-Native Companies0.99
    6. EURO1.00
    7. WIZ'0.79
    8. DRATA0.96
    9. PROCORE1.00
    10. S ARR (in M)0.98
    11. LEGORAdeel.SIERRA0.98
    12. Harvey gleonV0.98
    13. workday1.00
    14. /:naplan0.98
    15. servicenow1.00
    16. $1001.00
    17. cohere0.99
    18. ope0.88
    19. $800.94
    20. $600.94
    21. $401.00
    22. AIEng0.93
    23. $200.99
    24. EURO0.99
    25. SO0.66
    26. 00.99
    27. 11.00
    28. 21.00
    29. 41.00
    30. 51.00
    31. 61.00
    32. Years from 1 to 100M ARR0.98
    33. ICONIQ1.00
    34. er1.00
    35. Notes: Assumes 24 months from founding to $1M ARR if not disclosed. Assumes exponential growth from founding date to most recent year company reached $10OM ARR0.99
    36. Source: Historical filings and public company announcements0.98
    37. AlEngineer0.96
    38. AGENTS NEED MORE THAN A CHAT0.99
    39. EUROPE1.00
    40. PRESENTED BY1.00
    41. Google DeepMind0.98
    42. JACOB LAURITZEN/CTO0.99
    43. LEGORA1.00

  • 2:09 #19 done17 line(s)

    shot 19·sharpness 1671.4

    1. OPE0.97
    2. LEGORA1.00
    3. COLLABORATIVE AI FOR LAWYERS1.00
    4. Goal of vertical Al1.00
    5. epM1.00
    6. AIE1.00
    7. Make agents complete complex work e2e1.00
    8. neer1.00
    9. rko0.78
    10. AIE0.92
    11. AlEngineer0.98
    12. AGENTS NEED MORE THAN A CHAT1.00
    13. EUROPE1.00
    14. PRESENTED BY1.00
    15. Google DeepMind1.00
    16. JACOB LAURITZEN / CTO0.97
    17. LEGORA1.00

  • 2:18 #20 done20 line(s)

    shot 20·sharpness 1904.9

    1. LEGORA0.93
    2. COLLABORATIVE AI FOR LAWYERS0.99
    3. toge0.92
    4. AlEngine0.91
    5. enAl0.93
    6. Snorkel0.95
    7. Goal of vertical Al1.00
    8. gineer1.00
    9. AlEngineer0.98
    10. Microsoft1.00
    11. Make agents complete complex work e2e1.00
    12. Tessl1.00
    13. AlEngineer0.93
    14. AlEngineer0.98
    15. AGENTS NEED MORE THAN A CHAT1.00
    16. EUROPE1.00
    17. PRESENTED BY1.00
    18. Google DeepMind1.00
    19. JACOB LAURITZEN/ CTO0.98
    20. LEGORA1.00

  • 2:25 #21 done12 line(s)

    shot 21·sharpness 1515.8

    1. eo4j0.94
    2. Engineer0.99
    3. New Economics of Production1.00
    4. Tessl1.00
    5. Engineer1.00
    6. AlEngineer0.98
    7. AGENTS NEED MORE THAN A CHAT1.00
    8. EUROPE1.00
    9. PRESENTED BY0.99
    10. Google DeepMind1.00
    11. JACOB LAURITZEN/ CTO0.98
    12. LEGORA1.00

  • 2:34 #22 done32 line(s)

    shot 22·sharpness 1726.7

    1. AlEngineer0.97
    2. raintrust0.99
    3. AlEnginee0.98
    4. EUROPE1.00
    5. EUROPE0.92
    6. eer1.00
    7. :neo40.80
    8. AlEngine0.99
    9. -kel0.84
    10. AlEnginee0.98
    11. EUROPE0.99
    12. EUROPE0.95
    13. Planning1.00
    14. Doing1.00
    15. Reviewing1.00
    16. Tess0.99
    17. SENTF0.99
    18. work1.00
    19. work1.00
    20. work1.00
    21. AlEngineer1.00
    22. har0.94
    23. AlEnginee0.97
    24. EUROPE0.91
    25. EUROPE1.00
    26. AlEngineer0.98
    27. AGENTS NEED MORE THAN A CHAT1.00
    28. EUROPE1.00
    29. PRESENTED BY1.00
    30. Google DeepMind1.00
    31. JACOB LAURITZEN/ CTO0.98
    32. LEGORA1.00

  • 2:48 #23 done25 line(s)

    shot 23·sharpness 1735.2

    1. AlEngineer0.99
    2. raintrust0.99
    3. EUROPE0.99
    4. OpenA1.00
    5. ineer1.00
    6. AlEngineer0.99
    7. orkel1.00
    8. EUROPE1.00
    9. Planning1.00
    10. Doing1.00
    11. Reviewing1.00
    12. SENTRY0.85
    13. work1.00
    14. work1.00
    15. work1.00
    16. AlEngineer0.99
    17. onar1.00
    18. EUROPE1.00
    19. AlEngineer0.98
    20. AGENTS NEED MORE THAN A CHAT1.00
    21. EUROPE1.00
    22. PRESENTED BY1.00
    23. Google DeepMind1.00
    24. JACOB LAURITZEN/ CTO0.98
    25. LEGORA1.00

Transcript

211 cues· 2,576 words· 13,707 chars

  1. 0:16 How's everyone doing?
  2. 0:17 Still good?
  3. 0:19 Great.
  4. 0:19 It's 5 PM on a Friday.
  5. 0:21 There's just me and two more people behind you and Friday beer.
  6. 0:24 So I'll try to be a little bit quick here.
  7. 0:27 I'm here to talk to you guys today about vertical AI and complex agents and why I think they need more than just the chat.
  8. 0:36 If you've ever worked with a long-running complex agent, you've probably tried something like this.
  9. 0:41 Sorry that it's all white, I can see the flashbang in your guys' face.
  10. 0:45 You turn to research something, draft a contract, make no mistakes, and it starts thinking, it starts reading, launches a bunch of sub-agents, does web search, writes files, launches more sub-agents, does more reading, writes more files, keeps going, takes forever, after 30 minutes,
  11. 1:05 It gives you your contract.
  12. 1:06 You take a look.
  13. 1:08 Clause 3 doesn't look right.
  14. 1:10 Did you make a mistake here?
  15. 1:11 Could you look at another document?
  16. 1:15 You're absolutely right.
  17. 1:18 Then you see this compaction.
  18. 1:20 That's when you know you can give up.
  19. 1:21 It's going to forget everything.
  20. 1:23 It's in the context state.
  21. 1:26 Anyway, it continues.
  22. 1:27 It keeps on going.
  23. 1:28 And you get a new contract.
  24. 1:31 Was it only clause 3 that was changed?
  25. 1:34 Probably not.
  26. 1:35 And so you end up in this state.
  27. 1:38 Not the greatest experience.
  28. 1:41 My name is Jacob.
  29. 1:42 I'm the CTO of Legora.
  30. 1:44 We are a collaborative AI workspace for law firms.
  31. 1:46 So we're a vertical AI company.
  32. 1:48 We have more than 1,000 customers, more than 50 markets.
  33. 1:51 We've raised a bunch of money.
  34. 1:53 We're growing extremely fast.
  35. 1:56 I'm being told maybe the fastest in history.
  36. 2:01 We are also hiring engineers in London.
  37. 2:03 So in case anyone's interested and wants to be on this growth journey, please talk to me after my talk.
  38. 2:10 Our goal and the goal of most vertical AI companies is to make agents complete more and more complex work end-to-end.
  39. 2:20 Doing that has changed a lot in the past six to 12 months because there are new economics of production.
  40. 2:26 So it used to be if you wanted to complete end-to-end work that you would be focused on doing the work.
  41. 2:33 That would be sort of the main thing is actually just getting it done.
  42. 2:39 Today, things look a little bit different.
  43. 2:43 Because right now, planning work and reviewing work is the new bottleneck.
  44. 2:46 So doing the actual work is extremely cheap.
  45. 2:48 It's very easy to do.
  46. 2:50 But now you have to spend time planning.
  47. 2:52 You have to get the non-functional requirements.
  48. 2:54 You have to get the specs.
  49. 2:55 And you have to spend a lot of time reviewing the work.
  50. 2:58 And if anyone's reviewed big PRs on GitHub, it really sucks.

Open at this second