Jacob Lauritzen

cto at @WeAreLegora and hiring big time

[Jacob Lauritzen @jacsebl]

if your agent asks me to contact someone it better know who that is

[Jacob Lauritzen @jacsebl]

Counting days until they allow us to chrome cast to in-flight screens

[Jacob Lauritzen @jacsebl]

@n2parko Took ya long enough

[Jacob Lauritzen @jacsebl]

@FajrEMalik fucking awesome

[Jacob Lauritzen @jacsebl]

Wanna learn more about Legora, culture, and scaling fast? No? ok don’t watch this then

[Salesforce Ventures @SalesforceVC]

🎬 In honor of @WeAreLegora's recent Series E fundraise, our own @nkkall had the opportunity to sit down with CTO @jacsebl for a wide ranging conversation on the company and how they approach building. A few topics covered:

👉 Why Jacob decided to join @MaxJunestrand & team at Legora.

👉 How Jacob thinks about maintaining culture while blitzscaling.

👉 How Jacob decides what to build with AI.

👉 The long-term ambitions for Legora.

👉 Why Legora partnered with Salesforce Ventures.

"The job today is very different from the job a year ago. You have to build the machine. You have to build the whole system that builds the product, which means we are architects of the system. We codify our tastes into the system such that agents have to work within those guardrails."

🔗 Watch Nowi's full interview with Jacob, now live on our Youtube channel: https://t.co/y3UCSZDvwh

[Jacob Lauritzen @jacsebl]

@realbasilchatha @swyx of all the quotes u couldve picked this is how ill be remembered 😭😭

[Jacob Lauritzen @jacsebl]

@jarredsumner I don't think so. Right now there's a huge wave of vulns being discovered, but at some point it will plateau and we'll end up with a few super battle-hardened deps. I'd much rather use those than my hand-rolled uniquely vulnerably brotli

[Jacob Lauritzen @jacsebl]

@pierrecomputer Huge, congrats guys!

[Jacob Lauritzen @jacsebl]

@swyx brave new world

[Jacob Lauritzen @jacsebl]

Got fable to hand-write weights of a 100k param model to recognize real versus fake legal citations.

Basically one-shotted a full NN that’s better than training it

[Jacob Lauritzen @jacsebl]

Why train when astra/fable can just raw dog the weights directly

[Jacob Lauritzen @jacsebl]

As expected with more training, performance converges between hand-written weights and trained weights

[Jacob Lauritzen @jacsebl]

Awesome to chat @swyx and @realbasilchatha !

[Basil Chatha @realbasilchatha]

@WeAreLegora announced a $550M Series E a couple weeks ago, so @swyx and I sat down with @jacsebl (CTO, @WeAreLegora) to talk about the nuances of building AI agents for the legal industry.

We got into:
-how one model release took Legora's citation accuracy from 70% to 99% with zero work
-why Legora runs on Azure (early customers refused AWS / GCP)
-why law firms punish partners for codifying what they know
-why Legora post-trains open-weight models
and much, much, more…

Enjoy!

Timestamps
0:00 Cold open
0:40 Meet Jacob Lauritzen, CTO of Legora
1:05 Starting out as underdogs in a Stockholm basement
1:54 How Legora grew from ChatGPT with Swedish law into a legal agent
4:12 The switch from assistant to agent
5:23 Which models Legora uses and why it added a model picker
7:48 Are today’s models good enough for lawyers?
8:20 What agents couldn’t do nine months ago
9:18 Testing the agent in simulated worlds and a 10,000-file data room
10:07 The evals team and Legora’s 100 legal engineers
11:03 When evals didn’t seem worth it and one release took citations from 70% to 99%
12:18 What legal practice could look like in five to ten years
14:18 Will the law firm partnership model survive?
16:40 What happens to paralegals and junior lawyers?
18:00 The billable hour and why earlier legal tech fell short
19:39 Where the money goes now
22:44 How Legora used its own product for acquisition diligence
23:59 How Swedish and American law firms differ
26:12 Product metrics when you can’t read the traces
27:41 Paraphrasing traces and narrowing failures down to 10 or 20 types
28:42 Why data privacy led Legora to run everything on Azure
30:04 Open-weight models and model arbitrage
30:55 Transcription models and why Legora doesn’t use ElevenLabs
33:33 The agent, Tabular Review, projects, and the Word add-in
34:24 How a lawyer uses Tabular Review
35:51 Why a spreadsheet works and what you learn from the first 10 rows
37:53 How the legal intake bot works in Slack
39:11 How agents create lists and lists start agents
40:15 Building products on those same pieces and acquiring Wexler
42:37 What forward-deployed engineers do at Legora
43:52 How legal engineers and forward-deployed engineers onboard a big client
46:43 What’s next for the forward-deployed team over the coming year
47:24 Why Legora resisted building APIs
49:26 Can the agent be the API?
50:12 The 300 Spartans slide and the plan to stay at 20 engineers
51:32 The bets nobody asked Legora to make
53:05 Post-training open-weight models for legal research
54:05 Search, retrieval, and late interaction
54:51 What makes a good legal agent harness and why drafting depends on search
56:25 Which model writes the best draft?
57:03 Hallucinations and Legora’s citation system
57:41 Do lawyers check the citations?
58:19 MapReduce or one big context window for a million-token contract?
1:00:16 Getting anonymized examples and the idea of buying a law firm for its data
1:01:53 Why Jacob advises against default embeddings
1:02:44 Will lawyers work in Legora or through Claude?
1:03:26 Will the model labs cut Legora off?
1:04:24 Law schools and scholarships for computer science majors
1:05:28 Hiring in New York, Stockholm, and London
1:06:09 Why most work eventually becomes legal work

Link in the comments.

[Jacob Lauritzen @jacsebl]

Post training was never the same

[Rikiya Takehi @rikiyatakehi]

We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.

We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: https://t.co/bOWCcFhCVq

[Jacob Lauritzen @jacsebl]

Runtimes should inject time budgets, not just token budgets. Give models an estimated time cost to potential actions and tool calls and see how it starts optimizing across both token and time efficiency.

[Tejas Bhakta @tejasybhakta]

models have no sense of time. they'll trigger a 30 min rebuild for a 2 line change instead of thinking ahead and batching the work

current frontier RL rewards token efficiency
future RL will reward time efficiency

the fast takeoff will feel absurd after this

[Jacob Lauritzen @jacsebl]

For extremely large document corpora with complex relations agents need to traverse at runtime, pre-computing the graph gave us huge latency and accuracy gains.

Today we’re releasing our latest version of legal research - now powered by a huge graph, or dare I say, ontology.

[Jacob Lauritzen @jacsebl]

@siggelabor I’ll be there brother

[Jacob Lauritzen @jacsebl]

i need a haircut i swear i look better irl. or someone send me some hats

[TIME @TIME]

Jacob Lauritzen describes how Legora's AI makes legal work quicker.

[Jacob Lauritzen @jacsebl]

@siggelabor actually u know what time of the year it is

[Jacob Lauritzen @jacsebl]

@imanradjavi Gone by IPO