Josef unveils AI engine to kill legal intake forms

Josef has launched a Rapid Ingestion Engine that uses AI to convert unstructured business inputs into structured data for legal workflows, while keeping document outputs rules-based and controlled.

The product targets a longstanding friction point in legal operations. Many legal teams have refined templates, clause libraries, and workflow logic over years. Yet intake often remains manual, relying on long forms and back-and-forth to translate commercial discussions into the fields those systems require.

Josef positions the new engine as a shift in where AI sits in the workflow. It argues automation gains are limited when AI is focused on drafting and other outputs, rather than extracting and validating the inputs that feed an established template process.

"Most teams are applying AI at the wrong point in the workflow - the output," said Tom Dreyfus, CEO and Co-Founder of Josef.

"The right place is ingestion. AI should read the mess that arrives before the template runs. The output should remain deterministic, controlled, and reliable. That's the death of the intake form," Dreyfus said.

How it works

The Rapid Ingestion Engine accepts unstructured material that typically arrives from the business, including email threads, meeting notes, term sheets, and draft summaries. It extracts key commercial terms and converts them into structured workflow data. It also flags gaps and prompts users for missing information.

The system then routes approvals based on workflow logic owned by legal teams and generates outputs using templates they have approved. Josef describes these outputs as deterministic and controlled, rather than open-ended text generation.

In practical terms, the product is designed to reduce re-keying and interpretation when commercial teams pass information to legal. It also aims to preserve the standardisation legal teams have built into their processes, including mandatory positions, pre-approved language, and required approvals.

Business intake

Josef outlined a scenario in which a marketing team negotiates a sponsorship arrangement over email, agreeing pricing, branding rights, and deliverables in a thread. In a conventional setup, the email chain is forwarded to legal and sits in a queue, or the marketing lead is sent an intake form with dozens of fields covering information already contained in the emails.

Under the new approach, a user uploads the email thread. The engine extracts the terms and maps them into a sponsorship agreement workflow, identifies missing fields, and asks for additional details where required. The user reviews and confirms the extracted information, and a draft agreement is produced from the legal team's approved template.

This focus on ingestion reflects broader changes in how internal users engage with legal services, particularly in organisations that have invested in self-service workflows and standard contract processes. Legal operations teams have spent the past decade building playbooks, standard templates, and workflow tools that depend on structured inputs. The gap has often been at the start of the process, where information arrives in emails, notes, screenshots, and informal documents.

User expectations

Josef's leadership framed the launch as a response to changing user experience expectations inside businesses. Internal clients increasingly expect legal workflows to match the ease of consumer applications. That has put pressure on legal teams to minimise form-filling and repeated data entry, without loosening standards or creating risk through uncontrolled drafting.

Sam Flynn, COO and Co-founder of Josef, said the company is focused on reducing friction at intake while maintaining legal oversight of process and content.

"UX expectations have changed," Flynn said.

"People don't want to retype information they already have. We're using AI to make ingestion effortless while keeping outputs deterministic - so the business experience feels modern without Legal giving up control," he said.

The announcement comes amid growing debate over where generative AI fits in legal work. Some organisations have experimented with AI-assisted drafting and review, while others have limited use to research and summarisation. A common concern has been the reliability and auditability of outputs, particularly where contracts and approvals must follow established policy.

Josef's Rapid Ingestion Engine takes a different route. It uses AI as a front-end interpreter of messy business communications, while keeping the final output bound to pre-approved templates and workflow rules. Josef links this approach to the idea that intake forms, as the main way to capture contract requirements, will play a smaller role as unstructured inputs become machine-readable.

It can also ingest common business artefacts such as email threads and meeting notes, and prompt users when details are missing. It is aimed at legal teams that already run structured workflows and want a lighter-touch experience for internal clients.