Startups

Legora Embeds OpenAI Models Into Agentic Legal Workflows

Stockholm-based legal AI startup Legora positions itself as an operating system for AI-driven contract review and legal work, powered by OpenAI's model stack.

Last verified:

According to the OpenAI Blog, Legora, a Stockholm-based startup founded in 2023, is using OpenAI’s models to automate time-intensive legal work—from contract review to large-scale document analysis. The company’s core insight is that legal teams are chronically under-resourced: junior lawyers face backlogs of contracts to review, and routine legal work often goes unfinished because of capacity constraints. Legora’s platform is designed to handle these repetitive tasks at scale, with CEO and co-founder Max Junestrand framing the ambition as making “every piece of legal work in the world” run on his platform.

How Legora Leverages OpenAI’s Model Stack

Legora began building use cases around GPT-3.5, the moment when many of its legal workflows became technically feasible, according to the OpenAI Blog. Today, OpenAI serves as a key model provider powering Legora’s agentic legal work and document review capabilities. The startup does not limit OpenAI’s role to the product layer; Legora’s internal teams—engineering, legal, and operations—also rely on OpenAI’s tools including ChatGPT Work and Codex for their own workflows. This dual deployment creates a feedback loop: advances in OpenAI’s model capabilities are translated directly into new product features for Legora’s customers.

Agentic Debugging and Prototyping Accelerate Internal Development

Legora’s use of AI agents internally demonstrates a distinct productivity pattern. According to Jacob Lauritzen, CTO at Legora, coding agents have compressed the engineering cycle: instead of committing to a single solution path, teams now “explore 20 solutions at the same time and, within days, pick the one we know will be the best.” The acceleration is most visible in debugging. Legora reports that bug diagnosis and repair—a task that previously required multiple engineers investigating for hours—now takes approximately five minutes, with coding agents searching millions of log lines and pursuing parallel explanations simultaneously.

Legora’s own legal department experiences the productivity gains the product is designed to deliver. Emma Kadri Bergström, General Counsel at Legora, oversees an NDA playbook built by analyzing the company’s historical agreements. The playbook allows Legora to automate first-pass review of new NDAs, freeing the legal team to focus on negotiations and non-standard terms. This mirrors the core value proposition for customers: agentic systems handle routine triage and analysis, leaving humans to concentrate on judgment-heavy decisions.

Why This Matters

Legora’s business model tests whether large language models can fundamentally reshape legal operations. If agentic workflows successfully absorb the routine work that currently consumes junior lawyers—contract review, document classification, playbook creation—then legal departments may redirect budget from hiring toward AI infrastructure subscriptions. For OpenAI, the Legora case study demonstrates model-as-a-platform traction: a customer building both product and internal operations atop OpenAI’s models creates deeper technical lock-in and feedback loops that inform model development. The short bug-fix cycle (measured in minutes, not hours) is particularly significant because it indicates that agentic systems are now fast enough to compete with human debugging workflows—a precondition for broader enterprise adoption in high-stakes domains like law.

Frequently Asked Questions

What does Legora's platform do?

Legora builds agentic workflows that automate legal work—including contract review, NDA analysis, and large-scale document processing—using OpenAI's models to handle tasks that traditionally require junior lawyers to work overnight.

How is Legora using AI internally?

Legora uses OpenAI's coding agents and ChatGPT Work internally. Engineering teams can prototype multiple solutions in parallel, and debugging is accelerated; the company reports reducing bug diagnosis from hours to approximately five minutes.

Why did Legora choose OpenAI?

According to the OpenAI Blog, Legora has relied on OpenAI's models since GPT-3.5, which unlocked the legal use cases the founding team had been exploring. OpenAI remains a key model provider for Legora's core workflows.

#agentic-ai #legal-tech #openai #generative-ai #workflow-automation