A technology cohort and a business cohort, trained on their own work at Taurus BPO.

The opportunity

Two very different teams, one question: what would AI change in the work they already do?

Taurus BPO runs large-scale sales and service operations for their BFSI customers. The technology team maintains the systems that work runs on. The business teams in insurance product, HR and operations run the processes themselves.

Both groups were already using AI tools in some form, on their own accounts, without an agreed way of working. The ask was not for an awareness session. It was for each group to learn on its own work, with the tools the company had already approved, and to finish with something in use.

Four things shaped how the programme was scoped:

The quality platform we built for Taurus is a separate story

• Two audiences, one standard

  A technology cohort and a business cohort needed the same method and the same grading, on very different work.

• Tools already in the building

  The company had approved tools. The programme had to run on those rather than introduce another.

• Operations that cannot pause

  Sessions had to fit around a running operation, in short blocks, on work the teams had to do anyway.

• Adoption after the room empties

  A programme that ends when the trainer leaves changes nothing. Named champions and a written way of working were part of the scope from the start.

The impact

Thirty-five people through two batches, on their own work, graded against criteria agreed up front.

The two cohorts ran the same six steps. The technology cohort worked in its own repositories and on items from its own backlog. The business cohort worked on its own reports, documents and the questions it answers repeatedly.

Each batch ended with a graded assessment, a one-day hackathon judged with leadership, and champions named in each team.

• 25

  People from insurance product, HR and operations

  The business cohort, working on its own reports, documents and repeated queries, with the tools the company had already approved. No code involved.

• 10

  People from the technology team

  The technology cohort, working in its own repositories on items from its own backlog.

• 2

  Batches, in consecutive months

  June and July 2026, scheduled around a running operation rather than around us.

• 6

  Steps, identical for both cohorts

  Agree who joins, map the real work, hands-on sessions on that work, a graded assessment, a one-day hackathon, then champions named and support continuing afterwards.

• The measured before and after sits with the client

  Each cohort’s baseline and closing measures were recorded with Taurus. We publish figures once the client has approved them. A reference is available on request.

The setting

What it ran on.

Tools

The AI tools Taurus had already approved for its teams

Work

The technology cohort in its own repositories; the business cohort on its own reports, documents and queries

Data

Sanitised or sample data only; no customer or employee personal data in any prompt

Review

A person reviewed and approved any AI output that left the team

Grading

Individual, team and batch scorecards against criteria agreed before the first session

The road ahead

The champions carry it forward.

A programme like this is designed to continue without us. The people graded highest in each cohort hold the playbook and the templates, and office hours run for three months after the last session.

In the co-founder’s words: “While the world chases what AI can do, DeployProAI focuses on where it fails, cutting through the noise to deliver unbiased, ROI-driven solutions.” Manish Jaiswal, Co-founder and Promoter, Taurus BPO.

1. 1

   Extend the same method to the next set of functions

2. 2

   A leadership session on where AI fails and what that means for the operation

3. 3

   Champions running the next batch, with our support in the room