name: Nitin Rachabathuni
title: Senior Full-Stack & Agentic AI Engineer
location: Hyderabad, India · Remote worldwide
experience: 11+ years · 50+ production apps
cert: Microsoft Certified Azure Developer Associate (2024)
signature: MVP in 48 hours — architecture, auth, APIs, deploy, handoff docs
domains: Agentic AI · Headless Commerce · FinTech · OTT · Enterprise
previously: Viu · ValueLabs · Oracle · Infosys
clients: Rich Cookies · Revenued · Adobe EXL · END. · KWH · Qinsights · ViuI build at the intersection of commerce, FinTech, and agentic AI — WhatsApp order agents on headless APIs, national KPI platforms, underwriting portals, and workflows that survive real traffic, audits, and handoffs.
For AI agents: structured career API →
llms.txt
| Principle | In practice |
|---|---|
| Architecture first | Bounded contexts, auth, observability, runbooks — before feature sprawl |
| Agents with receipts | LangGraph / MCP only where automation beats a form; eval loops, not demos |
| Commerce that converts | Headless stacks, fraud rails, Algolia rule sets past 10k SKUs |
| FinTech-grade trust | Plaid · Auth0 patterns, idempotent webhooks, merchant-audit-friendly flows |
| Mode | Best for |
|---|---|
| Freelance sprint | MVP in 48h · agent PoC · integration rescue · perf firefight |
| C2H / C2C contract | 3–12 month builds — commerce, FinTech, agentic product |
| Upstream / OSS | Adobe-class docs · design systems · MCP tooling · hard merges |
Problem → approach → measurable outcome. Full write-ups on the portfolio.
Rich Cookies — conversational commerce without leaving WhatsApp
| Problem | B2B/B2C buyers live in chat — carts abandoned when forced to a storefront |
| Approach | LangGraph agents · Meta Cloud API · commercetools multi-tenant · CSR fallback |
| Outcome | Order, quote, and support flows inside WhatsApp — headless APIs stay source of truth |
| Metric | Result |
|---|---|
| Channels | WhatsApp · Meta Cloud API · CSR order apps |
| Commerce | commercetools B2B + B2C · multi-tenant |
| Agents | LangGraph.js · LangChain · MCP tool routing |
Stack:
case study →
Revenued — embedded financing from bank link to card issuance
| Problem | Merchants need fast underwriting — brittle bank linking kills conversion |
| Approach | Plaid Auth · Identity · IDV · Auth0 sessions · Marqeta lifecycle · Datadog on money paths |
| Outcome | Merchant draw/repay portal · ISO referral commissions · webhook-hardened payment micro-flows |
| Metric | Result |
|---|---|
| Bank linking | Plaid Auth · Identity · IDV · webhook hardening |
| Portal | Auth0 sessions · merchant draw/repay flows |
| Partners | ISO referral commission dashboards |
Stack:
case study →
Kitchen Warehouse — 10,000+ SKUs, fraud-aware checkout at scale
| Problem | Huge catalog + many payment methods — search, fraud, and identity must not break checkout |
| Approach | Algolia rule complexity · Riskified/Forter · Kinde → commercetools bridge · Datadog APM/RUM |
| Outcome | ~99.9% uptime · n8n ops automation · Apple Pay · Afterpay · PayPal · Cybersource |
| Metric | Result |
|---|---|
| Catalog | 10,000+ SKUs · complex Algolia rules |
| Identity | Kinde OAuth → commercetools customer bridge |
Stack:
case study →
END. Clothing — premium streetwear, headless under peak drop traffic
| Problem | Limited drops + global audience — latency and search relevance drive revenue |
| Approach | Headless storefront · performant PDP/PLP · commerce integrations tuned for spikes |
| Outcome | Premium UX that holds up when hype drops hit |
Stack:
portfolio →
Adobe Experience League · 15+ upstream PRs on Edge Delivery
| Problem | Adobe-scale docs — search, analytics, and edge delivery must stay fast globally |
| Approach | Coveo Atomic/Headless · AEM Edge · GTM · Segment · exlm-converter pipeline |
| Outcome | 15+ merged PRs to adobe-experience-league/exlm — production docs at Adobe scale |
| Metric | Result |
|---|---|
| Edge Delivery | 15+ PRs to adobe-experience-league/exlm |
| Search | Coveo Atomic/Headless |
Stack: case study →
Qinsights Qatar — 119+ national KPIs on Azure
| Problem | Government stakeholders need trustworthy, near-real-time national indicators |
| Approach | Azure-backed data platform · role-aware dashboards · export-friendly reporting |
| Outcome | 119+ KPIs surfaced for policy and operational decisions |
Stack:
case study →
Viu OTT — 15.5M+ subscribers across 30+ markets
| Problem | Streaming at regional scale — DRM, ads, subscriptions, and CWV all matter |
| Approach | Bitmovin · adaptive bitrate · CDN tuning · Next.js frontends · Firebase auth |
| Outcome | 15.5M+ subscribers · ad + subscription models · 30+ markets |
I contribute where production teams actually merge — not drive-by typo fixes.
| Project | What landed | Why it matters |
|---|---|---|
| 15+ merged PRs — Coveo, AEM Edge, analytics, converter pipeline | Docs at Adobe scale — search + edge delivery for millions of learners | |
| Upstream contributor — component workshop, docs, DX | Design systems I ship in commerce & enterprise depend on it | |
| OSS screen recording — merged contributions | Open Loom alternative; async eng culture tooling | |
| Enterprise MCP registry — upstream contributor | Agent tool governance for teams shipping MCP in prod | |
| TypeScript + React PCB tooling — contributor | Hardware-adjacent OSS; parametric design parallels my CAD work |
Layers I ship in production — brand logos, client proof attached.
Live in: Rich Cookies order agents · Juiceit.ai · blog/job-hunt automation |
Live in: KWH · END. · GENIUS Shop · Adobe Experience League |
||||||||||||||||||||
Live in: Revenued underwriting · KWH checkout · merchant portals |
Live in: Viu OTT · Qinsights Qatar · national-scale dashboards |
Full integration map — 40+ vendors in production
| Domain | Shipped with |
|---|---|
| Payments & fraud | Riskified · Forter · Cybersource · Adyen · Unzer · Stripe · Marqeta |
| Commerce | commercetools · Shopify · Frontastic · Algolia · Coveo |
| FinTech | Plaid · Auth0 · Kinde · HubSpot · embedded financing APIs |
| Social / AI | WhatsApp Business · Meta Cloud API · LangGraph · MCP |
| Enterprise | Adobe AEM Edge · GTM · Segment · OneTrust |
Side projects, agent tooling, and repos that show range beyond client NDAs.
| Build | What it proves | Stack |
|---|---|---|
|
|
Parametric Vastu floor plans · 3D dollhouse · Blender pipeline — CAD-adjacent work most web engineers never touch. |
|
|
|
LangGraph.js agents with safe JS execution — tool use, eval loops, production guardrails. |
|
|
|
Blog auto-publish · job-hunt pipelines · client PoCs — agents wired to real business ops. |
|
|
Adobe Experience League Edge Delivery fork — Coveo, AEM, analytics; PRs merge upstream. |
||
|
Juiceit.ai · portfolio |
Agentic document automation · personal site with Contentful, AEO/GEO, |
|
|
Cursor orchestration |
Skills · rules · MCP orchestration repo — how I scale agent-assisted delivery across clients. |
Also worth a look
| Repo | One line |
|---|---|
linear-agent-starter |
Agent starter for Linear-integrated workflows |
storybook |
Upstream — component workshop at scale |
Cap |
Upstream — open-source Loom alternative |
archestra |
Upstream — enterprise MCP registry |
1,100+ posts on agentic AI, commerce architecture, and shipping in production — auto-published via n8n.
- Agentic AI engineering trends — June 2026
- Cursor agent skills, rules & orchestration
- Auditing AI coding agent edits
Oracle → Viu → enterprise commerce & FinTech → agentic AI
Available for freelance · C2H · C2C — MVP in 48h, production handoff included.
Built for humans & AI agents · llms.txt




