01 · cold open
Live role-play · ~60 sec

Admitted Student × Second-Year

DevanshHey Kevin, congrats on the Haas admit! What are you most excited about?

KevinHonestly, I'm pumped. Berkeley brand, amazing classmates, AI entrepreneurship with Omri…

DevanshNice. What about the part you're not excited about?

KevinRecruiting. It's brutal.

Devansh(to audience) Quick show of hands — who here found recruiting the worst part of their MBA?

KevinYeah. And if recruiting is broken for us, the top candidates? It's broken everywhere. That's why we built Trackly.

02

Trackly

The fastest agent to unlock job opportunities.

Kevin Astuhuaman · Devansh Shah · Jasmine Rattan · Aaron · UC Berkeley Haas MBA

03 · why now
What changed

Job search is going agent-driven.

People used to click through ten job boards. Now they open Claude or ChatGPT and expect an agent to do it for them — and to pull in their network.

Every tool in hiring was built for Boolean search boxes. That world is ending.

04 · why incumbents can't fix it
The structural conflict

LinkedIn and Indeed work for employers. We work for job seekers.

$17.8B

LinkedIn FY25 revenue — almost entirely from employers paying to post, promote, and throttle free listings.

$7.5B

Indeed (Recruit HR Tech) FY24 revenue — pay-per-click model makes more money when jobs stay up longer.

Together $25B+ in public-disclosed revenue from working for employers. They literally can't show everything for free.

Sources: Microsoft 10-K FY25 (year ended Jun 30, 2025) · Recruit Holdings FY24 Annual Report (year ended Mar 31, 2025).
05 · what we built
Three things, done well

Speed. Coverage. Network.

Speed
Jobs in minutes, not hours.
Median 52 min vs. ~7.6h on LinkedIn in head-to-head tests.
Coverage
Every job, not just the paid ones.
43% of senior roles at top tech never appear on LinkedIn in the first 24h.
Network
Warm intros, not cold apps.
Surfaces alumni and personal-network paths into every role we find.

Native apps for people who want a feed. MCP and CLI for people already living inside Claude and ChatGPT.

06 · how we win
The moat

Self-healing scrapers. Three months of cooking.

We've built an agent that builds agents — reverse-engineered across every major recruiting platform. New company in, working scraper out.

Not reproducible overnight.Thousands of iterations, dozens of format-specific patterns, live in production.
Incumbents won't copy this.They'd cannibalize their own paid-listings revenue.
It compounds.Every new format we crack makes the next one faster.
07 · proof
Pulled off the shelf

Users came to us. Then the career center did.

100+

weekly active users at Haas — half the class, fully organic, no launch and no marketing spend.

1 of 1

Berkeley Career Management Center is using our jobs data this week — and exploring paying for it.

Students ask us to give it to friends and partners at other schools. The product is doing the selling.

08 · market
Where this goes

MBA → all students → the agent-economy substrate.

Phase 1 · Beachhead
MBA career centers
Berkeley first, then top-20 MBA programs. Recurring revenue per cohort.
~$50M ARR ceiling
Phase 2 · Expansion
All universities
Handshake is vacating the high-end space. We fill the gap that career centers actually want.
~$500M ARR ceiling
Phase 3 · Platform
Programmatic agentic hiring
Employers pay us to deliver matched candidates fastest — inside Claude, ChatGPT, and beyond.
$25B+ incumbent pool to disrupt
09 · team
Why this team wins

The exact four people to build this.

Kevin Astuhuaman · CEO / Product
7+ years PM. Agentic workflows expert.
Shipped Trackly across web, iOS, macOS, CLI, and MCP in 3 months — solo. Lives inside Claude Code and Cursor. Obsessed with the job seeker.
Devansh Shah · Domain & GTM
Ex-PM at Indeed.
Built products inside the $25B+ recruiting category. Knows how LinkedIn and Indeed make money — and exactly where they can't follow us.
Jasmine Rattan · Engineering
Career software engineer. 5+ years shipping production systems.
Owns matching, scoring, and context. Has been building software her entire career — now building the layer that actually understands job seekers.
Aaron · Operations
Makes the machine run like clockwork.
Runs user research, growth ops, partnerships, and everything that keeps the other three shipping. The reason we move at the speed we do.

All four are Haas MBAs. We're the users. We're not going back.

10 · ask
What we're raising

$250K SAFE · $20M cap.

Milestone · 3 months
5 university career centers actively using Trackly.
Berkeley + 4 UC / top MBA partners.
Milestone · 3 months
Thousands of weekly active users.
Scale the word-of-mouth loop beyond Haas.
Milestone · 3 months
~$1M ARR trajectory.
Career-center contracts at ~$200K ACV each.
What funds it
Coverage + speed engineering.
Two hires and compute to make the moat unquestionable.