Trending storyDeveloping

Legora runs 2,000 parallel legal agent evals on Modal Sandboxes

2 articles2 sourcessince Oct 26Last article 1h ago · Nov 2, 5:54 PM ET

Overview

AISummary of 2 articles

Legora runs 2,000 agent evaluations in parallel on Modal Sandboxes, according to a Modal post on X. The evaluations test a legal agent that works for hours across thousands of documents.

Modal says each run boots the full application pinned to a specific commit, so engineers can measure how their changes affect performance in a realistic environment.

Legora CTO Jacob Lauritzen, in an interview with Latent Space, describes how the company builds and tests its agent without being able to read client work traces. He describes test environments modeled on customers' work and says a citation problem improved from 70% to 99% after one release.

Written by AI from the articles below · updated Nov 2, 6:05 PM ET

Check the sources:

Latest developmentNov 2, 5:54 PM ET

Legora CTO explains how to build legal agents without seeing client data

Developments

2 developments

Nov 2, 5:54 PM ET · 1 article
Legora CTO explains how to build legal agents without seeing client data
Latent Space: Legora CTO explains how to build legal agents without seeing client data
Oct 26, 11:17 AM ET · 1 article
Legora runs 2,000 parallel agent evals on Modal Sandboxes
Modal: Legora runs 2,000 parallel legal agent evals on Modal Sandboxes

Article timeline

The articles in this story. Times are ET.

Nov 2

5:54 PM
Latent SpaceBlog
Legora CTO explains how to build legal agents without seeing client data
AIJacob Lauritzen, CTO of Legora, discusses how the legal tech company builds and tests its agent when it cannot read the traces of client work. He describes test environments modeled on customers' work and a citation problem that improved from 70% to 99% after one release. The conversation covers Azure deployment, Tabular Review, forward-deployed engineering, and the future of law firm partnerships and junior lawyer roles.

Oct 26

11:17 AM
Modal@modalOfficial
Legora runs 2,000 parallel legal agent evals on Modal Sandboxes
AILegora runs 2,000 agent evaluations in parallel on Modal Sandboxes, testing a legal agent that works for hours across thousands of documents. Each run boots the full application pinned to a specific commit, so engineers can measure how their changes affect performance in a realistic environment.

Heat Trend

Current comparable heat: 10
Peak comparable heat: 10Nov 2, 7:00 PM ET
Change in past 24 hours: –

The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.