SmeetPatel
3+ years building production software. Led a 15-member engineering team. Shipped enterprise systems at Johnson & Johnson. Now going deep on LLMs, RAG pipelines, and agentic systems.
The engineer
behind the code
Started shipping code professionally at 18. Built and maintained production systems across the stack — databases to React UIs — and eventually led an engineering team of 15 engineers.
Did a contract stint at Johnson & Johnson (Fortune 50), working on enterprise-grade frontend at scale — 22,000+ users, strict compliance and deployment protocols.
Now intentionally specializing in AI Application Engineering— not because it's trendy, but because it's where the most interesting engineering problems live. Going deep on LLMs, RAG pipelines, agent architectures, and evals. Building real things, not following tutorials.
Quick highlights
Technical skills
Where I've worked
Lead Software Engineer
CurrentStellarmind
- Own technical architecture and delivery across all client engagements — system design, tech stack, deployment strategy end-to-end
- Lead and mentor a 15-member engineering team across frontend, backend, and DevOps
- Reduced average project delivery time by 50% via architecture templates, CI/CD pipelines, and AI-driven development
- Conduct code reviews, set coding standards, and drive sprint planning across all active projects
Full Stack Engineer (Contract)
Fortune 50Johnson & Johnson
- Contributed to enterprise-scale systems serving 22,000+ users within large production codebases
- Collaborated with UX designers, backend engineers, and PMs to ship UI features meeting enterprise accessibility and performance requirements
- Operated within enterprise security, compliance, and deployment protocols at Fortune 50 scale
Full Stack Developer
2 yrsStellarmind
- Built and shipped 12+ client-facing web applications end-to-end using React, Next.js, NestJS, Node.js, and AWS
- Engineered backend for QuickCam (patented camera app) — subscription lifecycle, payment processing, multi-tier referral rewards
- Built real-time driver-user communication for FlexxyDrive using WebSocket and Socket.IO with concurrent connection handling
- Managed full AWS infrastructure: EC2, S3 + CloudFront, Lambda, RDS — provisioning through production monitoring
- Integrated payment gateways, OAuth providers, and notification services across multiple projects
Frontend Developer Intern
First rolePosimyth Innovations
- Worked on a 25,000+ LOC production React codebase powering WdesignKit WordPress plugin (30k+ active users)
- Shipped bug fixes and feature enhancements to real end users — first exposure to large-scale production code
Things I've built
Production systems across AI, enterprise SaaS, mobility, and consumer apps. More AI projects in progress via my 100 Days challenge.
AI Compliance Platform
EnterpriseMulti-Tenant Enterprise SaaS
Compliance management platform serving providers across DPDP, SOC 2, ISO 27001 frameworks — with AI-powered evidence verification and intelligent questionnaire assistance.
- AI evidence verification that auto-validates compliance docs against protocol requirements — eliminating manual review
- AI-assisted questionnaire module with intelligent suggestions based on existing compliance documentation
- Multi-tenant architecture with tenant-isolated data, RBAC, and BYOC storage integration
- Row-level security policies with role-appropriate visibility for auditors, clients, and admins
Krushiratna
7k+ UsersAI-Powered Agritech Marketplace
Full-scale agricultural platform with multi-role access, AI chatbot, automated content moderation, and real-time crop price comparison across market yards.
- AI chatbot using RAG + Pinecone, intent classification, SQL agent over PostgreSQL, and guardrails (PII detection, content safety)
- AI content moderation pipeline classifying uploaded videos — only farming-related reels pass to the feed
- Multi-tenant backend with 4 distinct user roles: farmers, buyers, corporates, and content creators
- Livestock/equipment marketplace, crop price comparison, farming news aggregation, short-form video feed
FlexxyDrive
10k+ DownloadsNext-Gen Ride Sharing Platform
Uber-style ride-sharing platform with microservices architecture, real-time tracking, and live driver coordination via WebSocket.
- NestJS-based microservices with gRPC serving separate Rider and Driver domains
- Real-time tracking with Google Maps APIs and WebSocket for live driver-employee coordination
- Deployed on AWS EKS with Docker containerization
The Fabcrew
DubaiDubai Home Services Platform
Urban Company-style home services platform for Dubai — microservices, real-time tracking, subscription payments, and multi-role workflows.
- NestJS microservices with gRPC for customers, employees, and drivers
- PostgreSQL schemas via Prisma — bookings, subscriptions, ratings, real-time availability
- Real-time tracking via Google Maps APIs + WebSocket for live driver coordination
- Payment gateway supporting one-time and recurring subscriptions (weekly, biweekly, monthly)
- Deployed on AWS EC2 + S3 with Docker
QuickCam
PatentedPatented Camera Application
Backend engineering for a patented iOS/Android camera application — subscription lifecycle, automated payments, and a multi-tier referral rewards engine.
- Subscription lifecycle management with automated payment processing
- Multi-tier referral rewards engine with complex payout logic
- 1k+ downloads with stable production performance
100 Days
of AI
Building AI projects, documenting learnings, going deep on LLMs, RAG, agents, and evals. One project or deep dive per day — no skipping, no fluff. Every output is real and shipped.
99 days remaining
Let's build
something great
Open to AI engineering roles, interesting product companies, and conversations about building with LLMs. Drop a message — I respond promptly.
Send a message
Fill in your details — then choose WhatsApp or Email to send.