Thesis
The post-SaaS era: why the next generation of legal tech won't look like SaaS
Brazilian legal tech was built on SaaS logic: sell better software so lawyers do the same work with less friction. What changes when AI starts to do the lawyer's work?
The first decade of legal tech in Brazil solved management. SAJ, Projudi, Projuris, Aurum. Good software that makes the lawyer spend less time on deadline tracking, case organization and documentation. The promise was direct: same work, less friction. The next generation solves something else. It substitutes the work.
The structural limit of SaaS
SaaS is a tool. It's worth the time it saves the person operating it. If a legal management system cuts 30% of a lawyer's time on administrative tasks, it's an excellent deal for the firm. But the firm keeps the same headcount.
What happens when the tool doesn't save time — it replaces the function? That's the question vertical AI agents are forcing across every knowledge-intensive market, legal included.
What a real legal copilot must have
It's not LLM with a pretty wrapper. It's an operational layer plugged into SAJ/Projudi, reading citations in real time, generating briefs in the firm's style, learning local jurisprudence and monitoring deadlines with formal accountability.
The critical difference: accountability. The agent doesn't suggest — it delivers. The lawyer reviews, doesn't draft from scratch. The entire process reorganizes around that exchange.
Why traditional SaaS players won't solve this
Incumbents have incentives not to cannibalize the current model. New startups have the advantage of building agent-first, without the legacy of screens and permissions designed for humans.
That's why MVA launched LawPilot instead of trying to push AI features into a classic SaaS product. Its architecture has been different since day one.
Other posts
View allMarket · 7 min
Tech Due Diligence in the age of agents: what Brazilian investors should check in 2026
Brazilian PE and VC funds still treat Tech DD as a compliance checklist. When the target runs on AI agents, that checklist misses 70% of the relevant risks.
Read postPlaybook · 8 min
Anatomy of a multi-agent architecture for judicial auction monitoring with Claude + AWS
How Leiloa.ai processes 12 thousand judicial auction notices per month with an ensemble of agents that uses Claude for semantic extraction and AWS Lambda for orchestration.
Read post
