Skip to main content

FAQ

Questions worth answering before the call.

Person working at a laptop with a code editor glowing on a second monitor
  • Every engagement follows the same four stages: Assess, Prototype, Build, and Hand over. Assess takes about two weeks and ends with a target architecture and a measurable definition of success. Prototype validates feasibility against real data before we commit to a full build. Build is production engineering — tests, observability, CI/CD, and a rollback path. Hand over transfers the system, the documentation, and the reasoning behind it to your team.

  • Then we'll say so. The point of the first two stages is an honest answer, not a foregone conclusion — if the data doesn't support the case, or a simpler approach would work better, that's what you'll hear, before either of us commits to a Build stage that shouldn't happen.

  • Two things: seniority and scope of commitment. Every engagement is staffed by the engineers who actually write the code — there's no account-management layer between you and the person building the system. And we design for the day we leave, which means documentation and enablement are part of delivery, not a wind-down task at the end.

  • Advisory work is scoped and billed independently of any build engagement that might follow. The deliverable is a direct, written recommendation — including a recommendation not to build something — not a proposal for more work. If a build engagement makes sense afterward, that's a separate conversation and a separate decision.

  • Both. Startups usually come to us with a specific AI feature or platform component that needs to go from idea to something real, fast, without hiring a team they can't yet justify. Larger organizations usually come to us with a system that already carries the business and needs to be modernized without a risky rewrite. The four-stage model works the same way for either.

  • Fewer than a larger firm would, on purpose. Every engagement is staffed entirely by senior engineers, which is a real constraint on how many we can run well at the same time. If we don't have capacity for your timeline, we'll tell you that on the discovery call rather than overcommitting.

  • Assess and Prototype are typically scoped as fixed fees, since the goal of both stages is a specific, bounded deliverable. Build is scoped once the prototype has told us what the system actually needs — as either a fixed-fee project or a staffed engagement, depending on scope. We'll tell you which makes sense for your problem on the discovery call, not after it.

  • Yes. Every engagement is staffed entirely by senior engineers — there's no junior bench routed to client work. That's a structural constraint on how many engagements we take on at once, not a marketing claim.

  • Yes. Some engagements are a fully staffed Kadrix team; others are a defined scope of work — architecture, a specific feature, a migration — delivered alongside your own engineers. We agree on the shape of that during Assess, not after.

  • You own the system, the code, and the infrastructure. The "Hand over" stage exists specifically so your team can run and extend what we built without depending on us — documentation, runbooks, and architecture decisions are part of the deliverable, not an afterthought.

  • We stay available for a defined support window after Hand over, agreed on during Build. After that, your team runs the system independently — that's the point of the Hand over stage. Further work later is scoped as its own engagement, not an open-ended retainer.

  • We sign NDAs as a matter of course and don't publish client names, logos, or case studies without explicit, separate consent. Where you see project examples on this site, they're described as representative or illustrative for exactly that reason.

  • It depends on the problem, not a fixed template — but recent work has used TypeScript, React, Node.js, Python, and Go for application code; PostgreSQL, Redis, and Snowflake for data; Kafka and Airflow for pipelines; PyTorch and LangChain for applied ML; and Docker, Kubernetes, Terraform, and GitHub Actions for infrastructure and delivery.

  • Book a 30-minute discovery call. Bring the constraint you're working against — timeline, budget, an existing system, a specific model you're evaluating — and we'll tell you honestly whether we're the right team before either of us commits to more.

Still have questions?

Still have questions? Book a discovery call