Streamline AI, based in Burlingame, California, announced two new features on July 15, 2026, aimed at turning the legal department into a faster and better-connected partner for the rest of the company. This isn't about another chatbot — the core is how AI assistants securely access verified legal information and how employees can reach lawyers where they actually work.

What Streamline AI actually does

Streamline AI is a platform for so-called in-house legal teams — lawyers employed directly within companies, not in law firms. These people handle dozens of requests daily: sales needs a contract reviewed, HR needs advice on an employment agreement, marketing needs a campaign assessed. The platform receives these requests, prioritizes them, and automates routine work using a set of AI agents.

The company was founded in 2020 by Kathy Zhu, former deputy general counsel at DoorDash, and Julian Wimbush, a former product lead at Google. Today, according to the company, the platform is used by over 1,000 lawyers at companies like Gusto, 8x8, and Bloom Energy. Last July, the startup closed a $14 million funding round.

Slack Intake Agent: the end of hunting for the right form

The first new feature is an agent built directly into Slack — the corporate communication tool that many employees use all day. Instead of hunting for the right form or switching between apps, an employee simply writes in plain language what they need. The agent itself identifies the type of request, captures relevant context, asks follow-up questions for missing information, and creates a structured request directly in the Streamline system.

Meanwhile, lawyers continue working in their own environment, where existing approval processes, permissions, and governance rules apply. In other words: for employees, it's as simple as messaging a colleague on Slack, but behind the scenes, a properly registered and managed case is created.

MCP Connector: why this is the most interesting part

The second — and technically most interesting — new feature is the MCP Connector. MCP stands for Model Context Protocol. It's an open standard introduced by Anthropic (the creator of Claude) in late 2024, which during 2025 was also adopted by other major players including OpenAI, Google, and Microsoft. In simple terms: MCP is a kind of "universal connector" that allows language models to securely tap into corporate systems and work with current data instead of making up answers.

Thanks to this connector, Streamline makes verified legal context accessible directly within the AI assistants that companies already use — specifically Claude, ChatGPT, Gemini, Microsoft Copilot, and Glean. An employee can thus ask their favorite assistant in plain language about the status of a legal request, request approval, or ask for a report — without having to open another system or bother lawyers with a routine query.

Security and the European context: GDPR takes center stage

When it comes to legal data, security is paramount. Streamline states that the MCP Connector is built for enterprise environments: it authenticates users via the OAuth 2.1 standard, enforces the same permissions that apply within the platform, and — crucially — does not store or cache customer data. It routes requests in real time and logs every interaction in an audit trail.

For Czech and European companies, this detail is more important than it might seem. Legal documents contain some of the most sensitive corporate data of all. The model where an AI assistant reads data only in real time but does not store it on its servers, and where it respects user permissions, fits well with the requirements of GDPR and the upcoming AI Act of the European Union. Streamline itself declares GDPR compliance and SOC 2 Type II certification. However, that doesn't mean automatic peace of mind — every company in the EU must itself keep track of where exactly its data flows, because the AI assistants themselves (ChatGPT, Claude, Gemini) are operated by non-European companies.

A broader trend: AI stops being a separate tool

Streamline isn't the only one heading in this direction. 2025 and 2026 are defined by AI moving away from a standalone browser window and instead "dissolving" into the tools where people already work. Anthropic released dozens of MCP connectors throughout the year for tools ranging from Blender to Google Drive, Microsoft integrated agent features directly into Word and Excel, and Salesforce launched an AI agent in Slack. Streamline applies the same logic to a narrowly specialized area — corporate law.

The practical impact is clear: the less switching between apps, the faster a company operates. If a salesperson doesn't have to learn how to use a legal system and can just write a query in Slack or ask Copilot, the barrier to collaborating with the legal department drops. The question remains how willing companies will be to let AI assistants access their most sensitive legal data — and that's precisely why Streamline emphasizes the audit trail and the fact that data does not permanently leave its system.

Availability and Czech language

Streamline AI is currently aimed primarily at the US market, and the product is available in English; there is no official Czech localization. However, the AI assistants through which the connector operates — ChatGPT, Claude, Gemini, and Copilot — handle Czech very well, so queries in Czech would theoretically work. For Czech companies with an international legal team, this is more of a demonstration of a trend than a tool to deploy immediately. Streamline plans a live demo of both features at a webinar on July 22, 2026.