We build agentic AI and voice systems.
This is the commercial half of Novessa. The systems are real, the customers are real, and the revenue funds the research on the other half of this site.
Three things, done properly, not a menu.
Systems that get things done on their own
Software that carries out a multi-step job by itself. It checks things, takes an action, confirms the result, and recovers on its own when a step doesn't go as expected, instead of just stopping.
Answers from your own documents
A system that answers questions from your documents and data instantly, grounded in what you actually have on file, instead of guessing from general knowledge or making something up.
Real-time phone and voice systems
Phone and voice assistants that respond as quickly as a real conversation, without the dead air, cut-off words, or robotic pauses of a typical automated line.
Practising on the hard version of the problem.
The research side of Novessa is trying to model how one specific person thinks, well enough that their own family would recognise it. That problem has no benchmark and no shortcuts.
The same team builds what you are paying for here. The evaluation habits, the honesty about what does not work yet, and the refusal to ship a demo dressed up as a product all come from that harder problem, not from a sales process.
Upcoming and advisory positions.
Not open roles today. If one of these fits you, contact us and we will talk.
Applied research across agentic systems and evaluation. You will move between commercial builds and the research problem behind them, depending on the week.
Turning a working prototype into something that survives real traffic. Model selection, RAG pipeline tuning, and the unglamorous work of making outputs consistent.
Low-latency speech pipelines for commercial clients. Barge-in, turn detection, telephony integration, and the parts of a voice stack nobody notices until they break.
The core research question, shared with the research side of Novessa. Building a stable behavioural model out of source material that is sparse, messy and personal.
Total ownership of design, across both the research presences and the commercial work. Everything we make gets touched by someone relying on it. If that makes you more careful rather than less, we should talk.