You publish your rates. Why does nobody else?

Because a range invites a negotiation and a quote invites a meeting. We would rather you self-select. Across both research rounds, none of the 20 IT services firms we studied published a price — so if the number is wrong for you, you have saved a call, and if it is right, the first conversation starts at scope instead of at price.

You list 55 services. Can one firm really do all of that?

Not simultaneously, and we do not pretend otherwise. Each service is tagged with the team it needs: some are delivered by the principal alone, some with one or two specialists, and some need a dedicated team we assemble for the engagement. If you ask for something we cannot staff well right now, we will say so and tell you when we could.

Why is recruitment in the same catalogue as software engineering?

Because it is the same buyer. The CTO who hires us for an architecture review is, three months later, trying to hire two engineers and failing. We already know what good looks like and can technically screen in a way a generalist recruiter cannot. TEKsystems, Accenture and Damco all put talent inside their services menu for the same reason.

How do you handle international staffing without an overseas placement licence?

We employ the engineer in India and supply their services to you remotely. That is an export of services rather than emigration, so the Emigration Act 1983 registration and its ₹20 lakh bank guarantee do not apply. Same engineer, same work, a completely different regulatory position — and it is zero-rated under GST with our LUT.

Will you tell us not to build something?

Regularly. A meaningful share of assessments conclude that an off-the-shelf tool does the job, that the data is not ready, or that the process should be fixed before it is automated. That answer goes in the report, in writing.

Are you certified? ISO 27001, SOC 2?

TOGAF 9 today. ISO 27001 implementation is underway and we will publish the certificate when it is issued rather than before. SOC 2 follows only if a concrete US enterprise deal requires it — roughly 70–80% of its controls overlap with ISO 27001, so doing them in that order is cheaper.

How do international clients pay, and is there GST?

Service exports from India are zero-rated under a filed LUT, so international invoices carry 0% GST. Payment arrives via Skydo or Wise with an automatic FIRA for your records. Indian clients are billed in INR plus 18% GST.

You are based in India. How does that work across time zones?

IST is UTC+5:30, which overlaps UK mornings comfortably and US East Coast mornings at the edge. We commit to specific overlap hours in the contract rather than promising round-the-clock availability we would not keep.

Your AI pilot did not fail. It was never architected.

A prototype gets to choose its own conditions. It queries a clean subset of documents that somebody exported by hand. It runs as one identity with access to everything. Nobody has asked what it costs at a thousand requests an hour, because it has never seen a hundred. And no auditor has asked why it produced a particular answer, because no auditor knows it exists.

Production removes every one of those permissions at once. That is the moment most teams describe as the model underperforming. It is almost never the model.

The four walls a pilot hits

1. The permissions model

Retrieval systems are, functionally, a new query interface over your entire document estate. If the index does not carry the same access-control decisions as the source systems, the assistant will cheerfully quote a salary band to the wrong person. Retrofitting row- and document-level authorisation into an index built without it is not a patch; it is usually a rebuild, because the embedding and chunking strategy has to change too.

2. Data residency and lineage

The prototype called a hosted API from a laptop. The production system has to answer where each byte was processed, under which contract, and for how long it was retained. For an Indian company selling into the EU, that question arrives twice: once under GDPR for personal data and again under the DPDP Act at home. It is answerable. It is not answerable late, and it is expensive to answer after the architecture is fixed.

3. Unit economics at real volume

Token cost is the visible line and rarely the decisive one. The decisive ones are re-embedding cost every time the corpus changes, retrieval infrastructure that has to stay warm, the evaluation runs that tell you whether a prompt change made things worse, and the human review queue for the fraction of answers nobody will ship unchecked. A feature that is comfortably profitable at pilot volume can invert at production volume, and the inversion point is calculable in advance.

4. Explainability under challenge

Somebody will eventually dispute an output that mattered. If the system cannot reconstruct which sources it used, which version of the prompt was live, and which model answered, the only available response is an apology. Provenance is a design decision made at the beginning or a forensic project undertaken at the worst possible time.

Why this is an architecture problem specifically

Every one of those four walls is a cross-cutting constraint. None of them can be owned by the team building the feature, because each one spans identity, data platform, procurement, legal and finance. That is the definition of architectural work: the decisions that are expensive to reverse and that no single delivery team is positioned to make.

The cost of finding out in week two that the data is not ready is a two-week assessment. The cost of finding out in month six is the six months.

This is also why the sequencing matters more than the tooling. Gartner's August 2025 forecast — that forty per cent of enterprise applications will carry task-specific AI agents by the end of 2026, up from under five per cent — describes a lot of organisations about to make these four decisions simultaneously and under time pressure. The ones who write the constraints down first will spend less.

What to do instead, concretely

Write down the access-control model for the corpus before choosing a vector store. If two documents in the set have different audiences, that fact determines the architecture.
Model cost at three volumes: today, the twelve-month plan, and the plan being wrong by 5×. If the third case is fatal, you have found a design constraint rather than a budget problem.
Decide what evidence an answer must carry before the first answer is generated. Sources, prompt version, model version, timestamp. It costs almost nothing at the start.
Name the governance obligations that apply — EU AI Act risk class, DPDP consent basis, sector rules — and map each to a control someone owns.
Only then pick the models and the frameworks. That choice is the reversible one, and it is the one everybody makes first.

None of this is a reason to move slowly. It is roughly two to three weeks of work, it is cheaper than one wasted engineering month, and it frequently concludes that the right build is smaller than the one that was planned.