BETA · Launching on September 1. Stay tuned for updates.
BENGALURU, INDIA · · IST
--:--:----:--
SUBSCRIBE VIA SUBSTACK
--:--:--
BENGALURU

byshashank.BETA

Globally fluent · Asia-rooted

--:--:--
DUBAI
BLR--:--SIN--:--SHA--:--DXB--:--LDN--:--

№ 02Workhorse

The Human Layer

Putting AI inside a working, distributed team.

Real, anonymised lessons from AI rollouts inside distributed services teams in India. Trust calibration, context management, and why judgment — not prompts — is the scarce resource.

7 essaysThe Edition frontFiling since 2024Updated JUN 2026
Into the edition
On the cover№ 02 · JUN 2026
This desk, set as an edition — a lead, the curated rows, and the live wire.The Edition · № 02 · JUN 2026

The lead

the signature read · Teams

TeamsDeep · 9 min

A Management Problem in a Technology Costume

Most failed AI rollouts are management failures wearing a model's clothes.

By Shashank9 min read
Read the full essay →

Deep cuts

3 essays

Long-form, load-bearing essays. Block out the time — these reward it.

Context

Context Is the New Headcount

What you can load into the model's working memory is now a staffing decision, not a technical one.

Deep8 min

Field notes

1 notes

Observed close-up, written fast. Reporting from inside the thing as it happens.

Practice · Bengaluru

Judgment Over Prompts

Why the most useful AI skill isn't the one most job postings ask for.

Field note5 min

Quick reads

2 reads · under 7 min

One idea, in and out. For the gap between meetings.

In series

read in order

Field Notes from a Rollout

Dispatches from AI rollouts inside real services teams.

  1. A Management Problem in a Technology Costume

    Most failed AI rollouts are management failures wearing a model's clothes.

    Deep
    9 min
  2. Anonymising Lessons Without Losing Them

    How to write honestly about your team's AI mistakes without naming a single person.

    Quick
    5 min

Where this desk connects