Questions we get asked.

Answers

• How do you evaluate an LLM application before launch?

  Build a test set of real questions with known-correct answers before writing any feature code, run the application against it, and score both accuracy and faithfulness to source material against a defined threshold before shipping — the same discipline as automated testing for any other piece of software.

• How do you stop an AI agent from hallucinating in production?

  Ground every answer in retrieved source material and require a citation, add confidence scoring that routes uncertain cases to a person, and log every action for audit. The fix is almost never a better model — it's the engineering around it.

• How long does a custom software project take?

  A focused build typically runs three to six months from discovery to a live production release, with working software demonstrated every two weeks along the way. Larger platform work runs longer, but a genuinely usable first version should always arrive early, not months into the build.

• How much does it cost to build an AI agent for a business?

  Most agent projects fall between a focused three-month engagement and a six-month build, depending on how many systems it needs to touch. We typically start with a two-to-four-week scoping engagement, priced separately, that produces a fixed roadmap before any larger commitment.

• RAG or fine-tuning for a company knowledge base?

  RAG, for nearly every business use case. It stays current without retraining, and every answer can cite its source — which matters for trust. Fine-tuning is worth it mainly when you need consistent style or format, not factual recall, and increasingly the two work together rather than as a strict either/or.

• Should we build AI in-house or hire an agency?

  Build in-house if you already have ML engineers and the use case is core to your product long-term. Bring in a partner for specialised, non-permanent work — evaluation infrastructure, agent orchestration, or a first production deployment your team hasn't done before this one.

• What does it take to move an AI prototype into production?

  Usually three things a prototype doesn't have yet: real monitoring so you actually know the moment it starts getting things wrong, a maintenance plan for the models and dependencies it quietly depends on, and a defined incident process for the day something inevitably breaks in production.

• What should an AI readiness assessment cover?

  Four things, in this order: which use cases are actually worth pursuing given your real constraints, whether your data can genuinely support them today, a clear build, buy, or wait recommendation for each one, and a phased roadmap with success metrics defined before any code gets written.