Getting something impressive working takes a week. Getting it accurate on your own material, safe with your own data, fast enough at your busiest hour, and provable to an auditor takes the rest. This is the part we do, and these are the six pieces of it.
This is for you if
The six
Most projects use two or three of these together. A knowledge assistant for a bank, for example, is retrieval plus a private model plus the testing to prove it. They are separate pages because they are separate decisions with separate costs.
Answers drawn from your own content, with a citation to the exact page, and only what each person is allowed to see. For staff, for customers, or as the layer other AI systems call.
Chunking, hybrid search, rerankers, permission-aware retrieval →Open models running inside your network, your own cloud region, or a machine with no internet at all. Nothing is sent anywhere. Sized, installed and kept current.
Llama, Mistral, Qwen, vLLM, GPU sizing, air-gapped →Adapting a model to your industry's vocabulary, formats and house style, using work your team has already done. After we have tried the cheaper options and measured what they gave you.
LoRA, QLoRA, distillation, evaluation sets →Systems that decide what to do next and use your tools to do it, with hard limits on what they may touch, a person on anything consequential, and a record of every step.
MCP, tool calling, approval steps, injection testing →Features inside the software you already sell, in your codebase and under your name. Separation between customers, cost per account, and a switch for the buyers who need it off.
Multi-tenant isolation, cost per account, staged rollout →Proving the thing works and keeps working, including systems we did not build. A scored test set in your own pipeline, and a report your auditor will accept.
Scored test sets, slice-level accuracy, red teaming →Describe what you are trying to build and we will tell you which two or three of these it needs.
Where projects stall
We see the same five, in roughly this order. None of them is about the model, which is why picking a better one rarely helps.
Without a set of real cases and agreed answers, "is it good enough" becomes a matter of whose opinion is loudest, and the project cannot be signed off by anyone.
Where this is handledTesting and evaluation, and it is the first thing we build on every project.Scans, tables and drawings behave nothing like clean text. A specification table read as one line of numbers produces confident answers that are wrong.
Where this is handledRAG and knowledge assistants, in the reading and splitting work.The demonstration used a public service and nobody asked where the data went until the security review. The project then waits for months.
Where this is handledPrivate LLM, and settling the data question before the build rather than after.Reading is easy. Writing a result back into an ERP, with the right permissions and an audit trail, is where the actual weeks go.
Where this is handledThe integration work, which is what we were doing for eleven years before this.Accuracy slid, the exception queue grew, and the team went back to the old way without telling anyone it had stopped working.
Where this is handledManaged operations and team adoption, on the workflow automation side.Our stack
A summary across all six. Each page carries the full list for its own area, with a plain explanation of what everything is for.
FAQ
Usually two or three together. Retrieval plus testing is the most common starting pair. Private is added when a rule requires it, and custom training is the last one to reach for rather than the first.
Yes, and it is a large part of this work. We start by measuring what is there, which often finds that the model is fine and the retrieval or the integration is not.
No, and we build so that you do not either. Everything talks to a single internal endpoint, so changing model or provider is a configuration change rather than a rebuild. That is a deliberate design decision from the first week.
Same practice, different problem. Workflow automation is about work your team repeats. This is about building the AI itself, into your systems or your product. Some projects need both and the two sides use the same engineers.
Next step
If there is a demonstration that never shipped, the reason it stopped tells us more than the demonstration does. If there is nothing yet, tell us what you want it to do and who would rely on it.
Ahmedabad, India. We work with teams in the US, UK, Europe, Singapore and the Gulf, and we are used to the time difference.