AI adoption practice · Bengaluru
You bought the AI licences. You didn't get the workflows.
Enterprises bought Coursera, Pluralsight and Copilot and saw no measurable P&L movement. The gap is workflow redesign, not tools. We install working agentic workflows and measure the workflow, not the logins.
Tools we help teams adopt
Chokmah is an AI adoption practice in Bengaluru. We install working agentic workflows inside mid-market and nano global capability centres and measure workflow throughput, not licence logins. MIT found 95% of enterprise GenAI pilots return no measurable P&L; the gap is workflow redesign, not tools.
- MIT found 95% of enterprise GenAI pilots deliver no measurable P&L return (MIT NANDA, 2025).
- Your real competitor is the unused Coursera, Pluralsight or Copilot licence you already pay for.
- The five-rung ladder runs from a free AI Reality Check to a monthly governance retainer, with prices published.
- We measure workflow throughput, not seats, and name the workflows you should not automate.
- Chokmah has shipped zero client engagements and says so; scenarios describe method, not delivered work.
MIT found 95% of enterprise generative AI pilots delivered no measurable P&L return.
The report locates the cause in an organisational workflow and learning gap, not model quality. The headline has been contested for its small base; we think it holds directionally, and we treat it as the reason to run a paid diagnostic before any build.
Your real competitor is the unused licence
Chokmah does not compete with the other training vendor. It competes with the Coursera, Pluralsight and Copilot seats you already pay for and nobody uses. Those seats were bought to change a workflow. The workflow never changed, so the seats sit idle, and the spend shows up as a line item with nothing behind it.
More training does not fix that, because the problem was never a skills gap. It was that no course can name the workflow that exists only inside your organisation. We start from the workflow (audit which one the licence was meant to change, instrument it, and rebuild it) rather than from another catalogue of content.
The evidence we walk in with
A METR randomised trial found experienced developers were 19% slower with AI while believing they were 20% faster.
METR has since said (February 2026) the result is historical and may not reflect current tools. We cite it with that caveat because the durable lesson is unchanged: measure the workflow, do not trust self-reported speed-ups.
What we sell: the ladder
Sold in order. We rarely start at the bottom. Rung 1, the paid diagnostic, is the wedge: it is what procurement actually wants and it disqualifies tyre-kickers.
| Rung | Offer | Shape | Price |
|---|---|---|---|
| 0 | AI Reality Check | 90 minutes, on-site, 15–25 people. They build one real thing from their own work, live. | Free |
| 1 | Adoption Diagnostic | 2 weeks. Interview 8–12 people, shadow 2 workflows, audit licence utilisation. Scored report. | ₹1.5–3L |
| 2 | Workflow Sprint | 4–6 weeks. One workflow, built with 3–5 of your people. You own the code. | ₹6–12L |
| 3 | Capability Cohort | 10–12 weeks, 20–30 people, capstone built on your own workflows. | ₹10–20L |
| 4 | Governance & CoE retainer | Monthly. Governance framework, agent evaluation harness, internal champion enablement. | ₹2–5L / month |
90 minutes, on-site, 15–25 people. They build one real thing from their own work, live.
2 weeks. Interview 8–12 people, shadow 2 workflows, audit licence utilisation. Scored report.
4–6 weeks. One workflow, built with 3–5 of your people. You own the code.
10–12 weeks, 20–30 people, capstone built on your own workflows.
Monthly. Governance framework, agent evaluation harness, internal champion enablement.
Prices are indicative opening hypotheses. Every engagement is quoted.
What makes this different
Five reasons a transformation owner picks us over a catalogue vendor or a big deck.
Chokmah versus a training catalogue
A course catalogue and a workflow install are not the same purchase. Here is where they diverge.
Free programmes teach these concepts, increasingly including orchestration and evaluation. What none can do is apply them to your workflow, your data and your governance.
Who this is for
We sell to the transformation owner, never the L&D head. Large Indian IT services and Fortune 500 MNC GCCs are not our customers. They do this in-house at a scale we cannot match.
Questions we get first
Not yet. Chokmah is a new practice and has shipped zero client engagements. Our scenario pages describe how we would run the work, and we label them as such. We would rather tell you that plainly than manufacture a case study you could disprove in one phone call.
The free AI Reality Check, then the paid two-week Adoption Diagnostic. We do not sell a cohort as a first engagement, and we do not skip to the bottom of the ladder.
No. Training is our delivery mechanism; the product is a workflow that runs faster or with fewer errors. It belongs in a transformation budget, not an L&D budget.
Transformation, never L&D. GCC headcount budgets are tightening; transformation spend is not. If we are pitching the L&D head, we are in the wrong meeting.
Yes. The five rungs run from free to ₹2–5L per month. In a quote-gated market where no India enterprise GenAI vendor publishes a price, publishing our ranges is itself the differentiator.
Start with a free AI Reality Check
Ninety minutes, on-site, with 15–25 of your people building one real thing from their own work. No slides. It exists to find out whether a paid diagnostic is worth either side's time.