Building the infrastructure of the agentic enterprise.
AI agents are moving from experimentation into production. The infrastructure needed to run them has to evolve with them.
Complexity should become infrastructure.
Teams should focus on creating valuable experiences with agents. Governance, execution, observability, security and resource management should be the job of the operating layer beneath them. When each team solves this alone, the organization pays for the same problem several times over and still ends up without a complete picture.
Governance, execution, observability, security and resource management should be the job of the operating layer beneath them.
Write once. Run any channel.
Voice. WhatsApp. Chat. Mobile. Email. API.
One agent. One runtime. One governance layer.
Tessera is the AgentOS: an operating system for artificial intelligence agents, built on its own kernel for mission-critical operations.
The channel changes. The agent doesn't.
Our principle is simple: the channel changes. The agent doesn't. Real-time voice, WhatsApp, chat and email run the same agent code, with the same business rules, tools and governance. Voice is not a separate product: it is part of the same runtime.
Designed and grounded in distributed computing, Tessera combines the actor model, massive parallelism, durable execution, checkpoints and replay. Matiq, our agent language with a compiler and typed effects, builds those guarantees into the code.
AI talks. Rules decide.
A deterministic engine keeps business policies out of the prompts. Identity, permissions, human approval, memory, audit and cost control are part of the architecture — not responsibilities left for each team to rebuild.
The platform also evaluates its own runs and proposes improvements to the agents, with tests, versioning and human oversight. Learning happens inside the same controls that govern the operation.
Running in the customer's own cloud, Tessera connects agents to enterprise systems and provides a common foundation to automate operations with scale, continuity and control.
One kernel. One language. The same agent on any channel.
Write once, run any channel.
The hard part isn't building agents.
Building an agent has become accessible. Running dozens of them, built by different teams in different frameworks, making decisions someone will audit later, is a different problem. That's why the platform was designed to govern what wasn't born on it too: an agent built in another framework stays where it is and points its model traffic, tool calls and invocations at Tessera. That is the path by which it gains an owner, a version, a policy and a trail. Nothing has to be rewritten to come in, and nothing is locked into the platform if it ever leaves.
Built by the people who built the systems it governs.
Felipe Scaphe founded Sciensa in 2010 and Tessera in 2026. The product and engineering core comes from Sciensa: sixteen years delivering mission-critical systems to banks and insurers, now building the platform that governs their agents. Sciensa is the team's track record, not a Tessera shareholder.
Who created Tessera and Matiq.
Felipe Scaphe is Tessera's founder and CEO. He created the platform and Matiq, the language agents are written in, and before that he founded Sciensa, in 2010.
Felipe Scaphe
Founder and CEO
Where it started.
I started in 2010 with a simple thesis: Latin America's hardest technology problems are in financial services, and solving them takes deep engineering, not offshore staffing.
Today, with the AI revolution, we're in a new paradigm. Access to a model no longer separates those who deliver from those who experiment: anyone has that. What separates them is the engineering around it, the layer that keeps an agent running in production, with policy, audit trail and cost under control.
2010
Sciensa is founded in São Paulo as a technology consultancy for financial services, growing out of SOA Expert, a school for enterprise integration and service-oriented architecture.
2016
Platform engineering practice expands; first corporate banking client.
2018
Data & AI practice launches; second office opens.
2020
200+ engineers; partnerships established with AWS, Google Cloud and Azure.
2024
Expansion into 7 industries, including Insurance. The AI Agents in production practice begins.
2026
Tessera is founded as the AgentOS for regulated operations.