Chaitanya Patil / AI workflow systems

I build AI operating systems for messy startup work.

At 20, I led five engineering interns inside a Y Combinator-backed startup to build an AI product from scratch. Since then, I have built sales, partner, onboarding, and content systems for international SaaS teams and multi-million-dollar founders.

View portfolio See projects

Career Trail

From startup intern to AI systems lead.

Soileum / Kointrack

Early startup operator

First reps in product context, startup execution, and turning loose problem statements into usable work.

Early AI adoption in content systems

Used AI to support content generation and repeatable production workflows inside an international live-events marketplace.

YC S22

AI product build lead

Chosen at 20 to lead five engineering interns at Peoplebox.ai, a YC-backed startup, and help build an AI product from scratch.

Sales, partner, and onboarding workflows

Built AI-assisted closing, affiliate enrichment, onboarding intelligence, and partner-page systems around founder-led SaaS work.

Pebble

Short-form content engine

Locked the brand kit, tone rules, and output rhythm into an AI content workflow capable of producing 20+ short-form pieces per week.

Portfolio

Work that survived contact with teams.

I work best where the goal is obvious but the operating context is scattered across founder memory, support transcripts, lead comments, CSV exports, product tools, campaign ideas, and half-built docs.

My job is to turn that pile into a system a team can run: clear inputs, prompts, review loops, handoffs, dashboards, landing pages, SOPs, and quality checks.

The pattern is consistent across Peoplebox.ai, Ticombo, Everis/Prosp, Pebble, and founder-led SaaS work: convert ambiguity into operating leverage, then leave behind a workflow that still works after the demo.

AI-assisted sales

Cut lead handling time by up to 80%.

Converted lead context, post context, comment behavior, and tone rules into human-reviewed DM drafts for faster closing work.

Before: 3 to 5 minutes per lead
After: About 1 minute per reviewed lead
Receipt: 56 personalized DMs sent in 12 minutes

Peoplebox.ai

Led five interns in a YC-backed AI build.

Ran team rhythm for a scratch-built AI product: goals, specs, agent loops, review lanes, QA context, and handoffs.

Selected at 20 to lead the build team

Partner enrichment

Moved top affiliates from export to relationship lane.

Audited a 614-affiliate Rewardful export, shortlisted the top 14, and built targeted outreach plus partner-page surfaces.

4 of 14 top affiliates moved into active 1:1 work

Partner pages

Built dedicated sales surfaces for high-leverage partners.

Created tailored landing pages, comparison content, referral-link routing, and partner handoffs for priority routes.

Demandii / LinkedOtter / FluidCRM / Lars Kruger lanes

Activation intelligence

Turned support friction into onboarding specs.

Mapped 30+ support conversations into email, Intercom tour, PostHog, and measurement specs for cleaner activation.

LinkedIn authentication identified as first wall

AI content ops

Built a 20+ pieces/week short-form system.

Locked prompt rules, scene structure, voiceover stages, caption rules, QA loops, and creator handoff docs around a brand kit.

Tuned for repeatable viral-format output

Projects

I build AI operating systems for messy startup work.

From startup intern to AI systems lead.

Work that survived contact with teams.

Cut lead handling time by up to 80%.

Led five interns in a YC-backed AI build.

Moved top affiliates from export to relationship lane.

Built dedicated sales surfaces for high-leverage partners.

Turned support friction into onboarding specs.

Built a 20+ pieces/week short-form system.

The private builds behind the public work.

Cloud Consciousness

Fall0ut

Anrxyst