📄readme.md ×

📊numbers.json ×

📜principles.txt ×

talha-usmani› readme.md · UTF-8 · LF ● synced

# readme.md # last updated · 2025.10 · toronto

available · product manager · 0→1 builder · tools-of-tools operator

I run product the way an operator would, by shipping the thing.

8+ years between systems thinking and a keyboard. Public-sector ops, Slack at scale, AI tools that actually save people work. I write code when the answer needs to ship today.

View projects ↵ Get in touch

// the short version const talha = { role: 'product manager', based: 'toronto · open to global', stack: ['figma', 'cursor', 'react', 'sql', 'python', 'mixpanel'], moves: ['ship', 'measure', 'instrument', 'cut'], openTo: true, };

02projects/

Three things I shipped in the last 18 months. Each one was built end-to-end — discovery, design, code, rollout. Click any card for the long write-up.

📘 case-atlas.md

● LIVE

Case Atlas — an AI casework assistant, built solo

A staff-facing tool that ingests 30k+ historical files, retrieves precedent in under a second, and drafts the response in the operator's voice. Saved my team about 6 hours per case.

Slack at the Ministry — 17 to 57 boards, 75% cost cut

Inherited a fragmented public-sector Slack, made it the spine of cross-org work. Cut tooling cost by 75% while expanding active boards from 17 to 57. Trained 200+ staff.

CRISPER — productizing a research method

Took a hand-rolled research framework from a single workshop to a 5-pillar product. Pricing, packaging, ICP, brand, the works. Now run independently by the founding team.

// two more in progress · write-ups coming soon

📘 internal-bot.md

DRAFTING

A Slack bot that triages incoming requests, tags by urgency, and routes to the right desk. Cut intake-to-action from 2 days to 4 hours. Write-up coming.

📘 ops-dashboard.md

DRAFTING

Built the leadership dashboard from scratch in two weeks: SQL pipelines, Metabase, weekly digest emails. Replaced a $40k/yr vendor. Write-up coming.

03numbers.json

A career in numbers, plotted as moving averages. Each spark is roughly 18 months of work.

Slack boards

57+40

Up from 17 when I inherited the workspace.

Tooling cost

−75%

Cut without losing a single workflow. Vendor consolidation + builds.

Cases indexed

30k

Historical archive, indexed and queryable in under a second.

Workshops led

Cross-functional. Research, ops, AI tooling, retros.

04principles.txt

Four operating rules I keep coming back to. Earned the hard way.

// rule 01

Ship the spike, not the spec

A working prototype settles ten meetings. If I can build it in a weekend, I'd rather have the artifact than the deck.

// rule 02

Measure the operator, not the metric

If the dashboard's up and the team is still tired, the dashboard is wrong. I instrument the work itself.

// rule 03

Cut before you add

Most product problems are subtraction problems wearing addition's clothes. I look for what to remove first.

// rule 04

The tool should speak the work

Software for operators is not the same as software for users. Use their nouns, their verbs, their tempo.

05contact.yaml

If you're hiring a PM who can prototype, I'd like to hear from you.

name: Talha Usmani email: talha@example.com linkedin: /in/talha-usmani github: /talha-usmani timezone: America/Toronto · UTC−5 availability: open · prefer remote-first teams # press ⌘K anywhere · or scroll the explorer ← # thanks for reading. ship something this week.