Most clients start with a 2–3 week AI opportunity assessment or a 6–10 week proof-of-value pilot. We agree the problem, the data, the success measure and the constraints in writing before any build begins.
In two-week sprints, each ending in a working demonstration, so you see progress continuously rather than only at the finish. Performance is measured against the criteria agreed at the start, including failure modes and edge cases.
You receive documentation, runbooks, source code and training for your team. Ongoing support is available only if you want it — there's no obligation to continue with us after handover.
Nepal overlaps almost entirely with most APAC working hours, so questions get answered same-day. For U.S. clients, Nepal is roughly ten hours ahead of Eastern time, so work moves forward overnight and is ready for review each morning.
Contracting & compliance
You can contract through AIB Holding Corp., a United States corporation, for a domestic counterparty and familiar contractual footing, or directly with Cognix Labs Pvt. Ltd. in Nepal — whichever suits your procurement process.
You do. Source code, models, weights and documentation are assigned to you on delivery. There is no proprietary runtime to keep licensing from us, and no lock-in.
Our delivery practices support Singapore's PDPA, MAS's FEAT principles and technology risk guidelines, IMDA's Model AI Governance Framework, U.S. federal and state privacy regimes, and the NIST AI Risk Management Framework. Nepal operations are governed by the Individual Privacy Act 2075.
Yes. Confidentiality and non-disclosure terms are agreed before any data is shared, and client security questionnaires and audit clauses are welcome.
Technical & deployment
Wherever you need it to: your cloud tenancy (AWS, Azure or Google Cloud), your private cloud, or entirely on-premise. Where data cannot leave the jurisdiction, we run models locally rather than calling an external service.
Least-privilege access, named individuals on every engagement, encrypted transfer and storage, secrets management, and no client data retained beyond what the engagement requires.
Every system comes with model documentation, evaluation records, and bias and failure-mode testing. Any decision affecting a customer's money, identity or access to a service has a human review point built in.
Pricing & team
Fees are quoted per engagement. Assessments and pilots are fixed price or fixed scope; dedicated pods and managed operations are billed monthly. Rate cards and client references are available on request.
Our Nepal engineering base lets us field experienced engineers at a cost that makes the second and third project easy to approve, without the churn typical of a commodity offshore vendor.
A named team of engineers, ML specialists and a technical lead who own your roadmap end to end — data, model, interface, deployment and handover — rather than work handed between departments.
Still have a question?
Tell us the process that costs you the most time or money today, and we'll tell you honestly whether AI is the right answer.