Aditya Nagpal
Aditya Nagpal
Author
The Lean Scale

How lean teams scale, function by function.

Which business functions move to AI agents today, which stay human, and how fast-growing companies sequence it. Written from inside 300+ teams building and scaling in India.

You'll only hear from me when there's something worth your time.

What's real today 2026
Finance & Accounting 60–99% manual work removed
Customer Support 49–90% tickets auto-resolved
L1 IT Service Desk ~90% routine requests handled
KYC Onboarding 40 sec down from ~6 weeks
300+ teams I watch build and scale up close: what they offshore first, what breaks, what compounds
Wisemonk Research primary studies plus function-level data on what agents actually do today
Who's writing this
Aditya Nagpal

Aditya Nagpal

Founder & CEO, Wisemonk · EOR for India

15 yrs building global teams
3 continents worked across
300+ global companies served
2,000+ employees on payroll

Writing from expertise built over a 15-year career in the industry + deep data-led research behind every word.

I have worked across three continents and sat in the boardrooms where billion-dollar decisions were made.

I started Wisemonk with one goal: to solve India hiring for global companies. Completely bootstrapped to a 40+ person team, it now runs employment for 300+ global companies and pays 2,000+ employees across India.

Running employment for hundreds of companies means I watch them build and scale up close: which functions they offshore first, where they hand work to AI agents, what compounds, and what quietly falls apart. I study that pattern across the whole portfolio.

So I decided to send what I'm seeing, function by function, straight to the founders and leaders scaling teams in India, at the right moment.

Published research

The research it's built on

Deep-dive reports from Wisemonk's own analyst work, free to read in full.

India Investment Intelligence 2026

Macro, capital flows, GCC ecosystem · 100+ institutional data points

India IT Services Analyst Report 2026

$297B → $315.4B trajectory · AI adoption · cost dynamics

India CX Market Statistics 2026

Four-layer CX stack · channel mix through 2030 · 20+ sources

Agentic offshoring, function by function

The series

  1. 1

    The ten functions you can offshore to agents, ranked by what's actually real today

    The full map, with the honest framing up front: autonomy is jagged. Tier 1 runs near-autonomous, judgment-heavy work stays human-led.

  2. 2

    Why finance and accounting is the most mature agentic frontier

    AP/AR and close agents hitting 60 to 99% manual-work reduction. High-volume, rule-bounded work is where agents already win, so start here and let it fund the rest.

  3. 3

    The customer support math has changed, and the floor just rose

    Resolution agents at 49 to 90%, time-to-resolve collapsing from 11 minutes to 2. The lesson, not the stat: over-automate and quality drops, so the hard cases route back to humans.

  4. 4

    The 90% number nobody at HQ believes: L1 IT service desks

    Triage agents auto-resolving most routine requests, plus the coding angle: ~26% more PRs, 10x migration speed. India's IT workforce isn't replaced, it's repositioned.

  5. 5

    What happens to a 700-person support team when one agent does two-thirds of the work

    Not headcount-zero, headcount-reshaped: fewer ticket-takers, more people on the hard 15%. Naming the fear directly.

  6. 6

    KYC that took six weeks now takes 40 seconds

    Onboarding agents for legal, compliance, and KYC, weeks to seconds, audit-grade compliance intact. Written as a practitioner, not a commentator.

  7. 7

    The recruiting function, rebuilt: 3x throughput, 85% faster time-to-hire

    Sourcing and screening agents, and the difference between commodity RPO and real talent-quality uplift. Recruiters move to hiring-manager advisory.

  8. 8

    Agents are negotiating with your suppliers now

    Procurement and source-to-pay agents closing deals with most tail-spend suppliers: 3% savings, extended terms, 4x ROI. The thing everyone assumed needed a human turns out not to.

  9. 9

    The unsexy truth: 80% of agent work is data plumbing, and 40% of projects fail without it

    The precondition no one sells: clean data and documented SOPs before any agent. Contrarian, honest, and the credibility play.

  10. 10

    How to sequence an agentic offshoring programme without lighting money on fire

    The Stage 1 to 3 roadmap: quick wins fund the programme, then scale into procurement, recruiting, and SOC, then orchestrate multi-agent workflows for 3 to 5x ROI.

  11. 11

    Jagged autonomy: the concept HQ needs before they greenlight any of this

    Some work is 99% autonomous, the work next to it is human-led. Miss the jaggedness and you either over-invest in pilots that fail or miss the Tier 1 wins.

Know which function moves next.

Function-by-function signal from inside 300+ teams building and scaling in India.

The Lean Scale Unsubscribe anytime