FAQ
The questions we get asked most
Grouped by what people are actually trying to work out: whether we can help, whether the AI will hold up, and whether your compliance team will approve it.
Working together
Engagements and commercials
What is the smallest sensible engagement?
A two-week AI Readiness Audit at a fixed price. You get a ranked automation backlog with effort, risk and projected savings against each item, and the fee is credited in full against a build started within 90 days.
Do you publish prices?
Yes, in INR and USD, on the pricing page. Fixed-scope packages can be bought directly by card or invoiced against your procurement process. Anything larger or unusual is scoped, and we say so rather than quote a number we would revise later.
How long until something is live?
Four weeks for one well-defined workflow in production, which is the Automation Sprint. Eight to ten weeks for a stalled pilot taken properly to production. Managed platform engagements start with a two-to-three week assessment before we sign any availability target.
Who owns the code?
You do, from the first commit, in your repository and your cloud account. Source, infrastructure-as-code and documentation are yours on payment. Exit assistance is contractual — a managed service that is hard to leave is not something we are willing to sell.
Can we pay by invoice or purchase order?
Yes. Card checkout exists for speed, not because it is the only route. NEFT, RTGS, wire transfer and PO-based procurement are all fine — request an invoice and we will work to your requirements.
AI, honestly
Accuracy, limits and what we decline
What happens when the AI gets something wrong?
It will, so the design assumes it. Every automation has a confidence threshold, an escalation path to a named human queue, and an audit trail. We agree an acceptable error rate with you up front and measure against it with an evaluation suite that runs in CI.
Will you tell us if AI is the wrong answer?
Yes, and roughly a third of our audits conclude exactly that — the highest-return action turns out to be fixing a broken handoff, replacing a manual export with an integration, or buying a product that already does the job. Those findings go at the front of the report.
Is there anything you refuse to build?
Yes, regardless of fee: autonomous decisions in regulated domains like credit or clinical care, surveillance of individuals, systems designed to deceive, dark patterns that make cancellation harder, and bulk unsolicited outreach. The full list is on our acceptable use policy.
Do you need our data to train a model?
Almost never. Most business automation is solved with retrieval over your existing documents plus a well-specified agent, not fine-tuning. We contract for zero retention and no training with every model provider we use, and we will say clearly if a case genuinely needs training data.
Which model do you use?
Whichever fits the task, cost and residency constraints — often more than one in the same product. We default to Claude for reasoning-heavy work and route simpler steps to smaller models. The architecture keeps the provider swappable, because pricing and capability move every few months.
Data and trust
Security, compliance and residency
Where does our data go?
For most engagements, nowhere near us — we deploy into your own cloud account under roles you grant and can revoke. Every third party that does touch data on our behalf is listed on our sub-processors page, which we keep current because clients rely on it for their own compliance.
Can you work with data that cannot leave India?
Yes. The practical pattern is keeping the authoritative record and audit trail in-region, minimising what crosses a boundary by redacting or tokenising identifiers first, and contracting for zero retention. Where even that is unacceptable, self-hosted open-weight models in your own VPC work — with a real capability trade-off we will measure on your task before you commit.
Are you SOC 2 or ISO 27001 certified?
Not yet. Both are in progress with controls implemented and evidence collected, but the audits have not concluded — so we do not claim certification. We can share control documentation and policies under NDA today. The compliance page states the position honestly, including target dates.
What service levels do you commit to?
On NoOps, 15-minute P1 response 24x7 and a 99.95% monthly availability target, with service credits applied automatically when we miss and monthly reporting with the raw incident data. Managed DevOps is business-hours. Full detail is on the service levels page.
Something not answered here?
The site assistant can look up services, pricing and client results and will tell you when it does not know. For anything commercial or contractual, ask a person — email [email protected] or book thirty minutes.
Let's find out what is actually automatable
Bring a process that annoys you. In 30 minutes we will tell you whether AI helps, what it would cost, and where it would fail — even if the answer is don't bother.
Or email [email protected] · we reply within 1 business day