AI Readiness Audit
We map your processes, data and systems, then tell you exactly where AI would pay and where it would waste your money.
Result: A clear, costed shortlist instead of a vague ambition.
12 · AI Consulting & Enablement
Most failed AI projects were the wrong project
The expensive mistake is not choosing the wrong model. It is spending six months automating something that was never the bottleneck. We start by finding where AI actually pays in your business, then build only that.
- Starting at
- ₹75,000
- Typical timeline
- 2–4 weeks for an audit
- Services in this line
- 7
The problem
Enthusiasm is not a strategy. The businesses getting real returns from AI picked two or three high-leverage processes and did them properly.
What we build
7 services in AI Consulting & Enablement
Each one is a page of its own — pick the closest match, or tell us your problem and we'll point you at the right one.
AI Strategy Roadmap
A phased 12-month plan — what to build, in what order, at what cost, with what expected return and what capability you need in-house.
Result: A plan you can fund and hold people to.
RAG Knowledge System
Your company brain — every document, SOP, contract and past project made searchable and answerable in plain language, with citations.
Result: Institutional memory that survives people leaving.
Custom LLM Fine-Tuning
Train a model on your own data, voice and domain when prompting alone is not enough — with an honest assessment of whether you actually need it.
Result: A model that speaks your business's language.
Prompt Engineering & Team Training
Hands-on training so your team uses AI properly, plus a prompt library and SOPs built around your actual work.
Result: Everyone gets better at it, with rules everyone knows.
AI Agent Monitoring & Maintenance
Ongoing observability, quality evaluation, cost tracking and model upgrades for AI systems already in production.
Result: You find out about quality drift before your customers do.
AI Governance & Policy
Internal AI usage policy, data handling rules and DPDP Act alignment so your team can use AI without creating legal exposure.
Result: Clear rules before an incident forces you to write them.
What you get
Included in every engagement
- Process-by-process audit with effort and impact scored
- A prioritised roadmap with real costs and timelines
- Honest assessment of what should not be automated
- Build/buy recommendation for each opportunity
- Data readiness assessment
- Executive summary your leadership can act on
How it's built
Typical stack
- Process analysis
- RAG architecture
- Fine-tuning pipelines
- Evaluation frameworks
- Monitoring and observability
We pick tools per project rather than forcing every client onto the same stack. If something in your business already works, we build around it instead of replacing it.
In depth
AI Consulting & Enablement: what you should know before you commit
The most expensive outcome available in this field is building the wrong thing well. A ₹6,00,000 system that works perfectly and addresses a process costing you ₹40,000 a year is a loss, regardless of how good the software is.
Why we sell assessment separately from building
Most businesses arrive with a specific idea — usually a chatbot — that came from a competitor's site or a LinkedIn post. Sometimes it is the right idea. Frequently the largest opportunity in that business is somewhere nobody has looked.
A vendor whose only revenue comes from builds has an obvious incentive to find a reason to quote. So we price the audit at ₹75,000, credit it in full against a subsequent build, and write the report to be useful to whoever reads it — including another vendor.
That structure is the whole point. It keeps our incentive on giving you an accurate map rather than an attractive proposal.
The seven engagements
| Engagement | Answers | Typical duration |
|---|---|---|
| AI readiness audit | Where would this pay, and where would it not | 2–4 weeks |
| AI strategy roadmap | In what order, at what cost, with what team | 4–6 weeks |
| RAG knowledge system | Make our documents answerable | 8–16 weeks |
| Custom LLM fine-tuning | Do we actually need training at all | 2 weeks assessment, then 6–12 |
| Prompt engineering & training | Get the team using this properly | 2–4 weeks |
| AI agent monitoring | Is what we built still working | Ongoing |
| AI governance & policy | What are the rules, and are we exposed | 2–4 weeks |
Three things we say more often than clients expect
You do not need fine-tuning. Almost every request for it is really a request for retrieval or better prompting. Fine-tuning teaches behaviour and format; it does not make a model know your documents. Clients want a model that knows their business, and training on their documents produces something that sounds like their documents while still being unable to state reliably what is in any of them.
Your data is not ready. Many SMB processes run on paper, WhatsApp and memory. Where nothing is recorded, most AI opportunities become available six to twelve months after you start recording something. That is a cheaper finding than a system with nothing to learn from.
Do not build that one. An audit finding twelve exciting opportunities and recommending all twelve is a sales document. We have completed audits whose main recommendation was to fix a spreadsheet and commission nothing.
Your team is already using AI, without rules
In almost every organisation we assess, staff are using AI tools that were never approved, on data that should not have left the building, with no record of it.
This is not misconduct — nobody told them not to, the tools are free and useful, and the boundary is genuinely unclear. But under the DPDP Act, pasting customer data into a third-party tool is a disclosure to that processor, and most free consumer tools have terms permitting training on inputs.
The fix is not prohibition, which drives usage onto personal devices where you have no visibility at all. It is a short set of rules people can follow, plus approved tools that make following them easy. That document takes two to four weeks and is increasingly what enterprise and international clients ask for during procurement.
Systems degrade quietly, which is why monitoring is a service
Conventional software fails loudly. An AI system that has degraded returns a well-formed answer that happens to be worse than it used to be — no error, no alert, no signal until complaints accumulate.
The causes are ordinary: a provider updated the model behind an endpoint, your product data changed, customers started asking about something new, someone edited a prompt without testing it.
The artefact that makes this manageable is a fixed evaluation set — a few hundred real inputs with known-good outputs, run regularly. It turns a model upgrade from a gamble into a decision, and it is the thing clients most often did not have before we started.
What these engagements cost
Readiness audit from ₹75,000, credited against a build. Strategy roadmap from ₹1,50,000. Team training from ₹1,00,000 for ten to twenty people. Governance policy from ₹1,25,000. Fine-tuning feasibility assessment ₹75,000, which has ended more than one engagement with a recommendation not to proceed.
RAG knowledge systems run eight to sixteen weeks from ₹3,50,000. Monitoring is ongoing from ₹40,000 a month, and we take on systems built by other vendors — a good share of that work is exactly that.
Every document produced is yours without restriction. Take it to another vendor, use it internally, or shelve it.
FAQ
Questions people actually ask
Will you tell us not to build something?
Regularly. Roughly half of what clients arrive wanting turns out to be better solved with a spreadsheet fix or a process change. Saying so is the job.
Do we have to build with you afterwards?
No. The audit and roadmap are standalone deliverables you own and can take to anyone.
How much data do we need for fine-tuning?
Usually a few hundred to a few thousand good examples. But most businesses that ask for fine-tuning actually need retrieval instead — cheaper, faster and easier to update. We will tell you which you need.
Other AI service lines
Next step
Let's work out what's worth building.
A 30-minute call. We'll tell you what we'd do, roughly what it costs, and whether it's worth doing at all.