Software we build, and you own
Plenty of what we are asked for has no model in it at all: a platform, an integration, a pipeline, a system that has to be right. We take that work, and we take it end to end.
You own what we build, source and all.
Twenty years of engineering, two US patents, and a decade building for US government, military and intelligence customers.
What we build
Data and analytics platforms
Your data in one place, modelled so it can answer questions, in front of the people who need it. Warehouses, pipelines, dashboards, and the unglamorous plumbing that decides whether any of it is trusted. Analytics projects usually fail on the ingest and the definitions rather than on the charts, so that is where the time goes.
Legacy modernization and replacement
The system that runs the business, was written a long time ago, and nobody wants to touch. Replaced or modernized incrementally, old and new side by side, with a way back at every step. A rewrite that goes dark for a year is a bet, and not one we ask a client to take.
Integration and automation
Making systems that were never designed to talk to each other work together, and taking the manual step out of the middle. Vendor APIs, file drops, message queues, the spreadsheet somebody rekeys every Tuesday. Often the highest return per hour on this page, because the work already exists and a person is doing it.
Custom applications
Internal tools, web applications, back-end services, data products, and the one-off system nobody sells off the shelf because it is specific to you. Delivered in small working increments you can see and steer, not a demo at the end.
AI platform implementation
Implementing or integrating somebody else's AI platform, or building the layer around it that makes it usable: retrieval, evaluation, guardrails, the connection to your data. You do not have to buy a model from us to get engineering from us. If an off-the-shelf model is the right answer, we will say so and wire it up properly.
Assessment and rescue
A second opinion on an architecture, a build that has stalled, or a vendor claim you want tested before you sign. We are comfortable telling you the honest answer is to do less, or nothing. For AI vendor claims specifically there is a fixed-fee service: independent AI evaluation.
How we work
Fixed-price steps, not an open meter. Each step has a price and a deliverable agreed before it starts. If the shape of the work changes, that is a conversation rather than an invoice you did not expect.
Small increments you can see. Working software early and often, so you are steering rather than waiting. The failure we are avoiding is the six-month build that turns out to have been aimed slightly wrong in month one.
You own the source. Not licensed, not rented, no runtime fee. It keeps working whether or not we do.
No lock-in, by construction. Keep us on to run and extend it, or take the source and the documentation and run it with your own team. Both are fine with us, and we would rather say so up front than have it discovered later.
We will tell you when the answer is no. Sometimes the right recommendation is a smaller system, an off-the-shelf product, or nothing at all. The public research on this site is a fair sample of how rigorously we test our own conclusions before standing behind them, and that is the same disposition you get on a project.
A short brief gets a real answer within one business day, not a discovery call. If there is enough to work with you get a scope sketch and a rough range back. If it is not a fit we will say so and point you at somebody better suited.
Where it runs is your decision, not a constraint of ours
Any of this can be built inside the boundary you answer for: on-premises, in your own cloud, or fully air-gapped, under HIPAA, CJIS, GDPR or export control. Our founder spent more than a decade building systems for US government, military and intelligence customers, where data sovereignty and reliability were the requirement rather than a feature.
That matters more than it used to. A great deal of modern tooling assumes it may phone home, and for some clients that assumption is the whole problem. Designing for the boundary from the start is far cheaper than retrofitting it after a review.
When the problem does need a model
Some problems are not a program anyone can write: reading messy documents, answering from twenty years of institutional knowledge, judging something an expert judges by eye. When that is the actual problem, we build custom AI models you own and run yourself, on your data and inside your boundary, on a platform we designed and built end to end.
Two specific versions of that work have their own pages: contract data extraction for legal teams, and independent evaluation of a vendor's AI claims before you sign for them. Every model we deliver also comes with a bill of materials describing what is in it.