About
I harness AI to build systems that think, adapt, and ship.
Founding AI Engineer at SalesMonk — I design autonomous agents, build context-engineered pipelines, and architect AI systems that most teams can't because they don't understand the models deeply enough. From durable workflow engines executing 100k+ actions/month to self-healing trading bots and voice agents, I turn raw model capabilities into production-grade intelligence.
Cluster Map
Next.jsNestJS
Express.js
TypeScript
Inngest
Redis
AWSPostHog
ReactReact NativeThree.jsFramer Motion
Tailwind CSSMarketing Sites
PostgreSQL
MongoDBPrisma
Qdrant
PythonExperience
SalesMonk AI
Next.jsTypeScriptNestJSPythonTemporal.ioRedisPostgreSQLPrismaAWS
VapiInngestSlack- Leading development of client-level memory & intelligence layer using mem0, Firecrawl, LlamaIndex & Qdrant
- Architected durable workflow execution engine using Temporal.io powering 100k+ multi-channel actions/month
- Built AI Agent using Vercel AI SDK and Parallel.ai for structured web research and lead enrichment
- Designed glassmorphic voice widget & multi-tenant outbound AI voice agents using Vapi
- Engineered real-time chat widget managing 2M+ visitors/day, increasing meetings booked by ~50%
- Led RAG-based LLM chat system with Slack integration, cutting first-response from minutes to seconds
- Designed account scoring pipeline processing 100K+ accounts using Inngest & Redis on serverless infra
- Developed shareable story SaaS product supporting 20+ media types, scaled to 50k+ stories
Evolv Fit
ReactReact Native
Node.jsExpress.jsMongoDBFirebaseAWS
GCPTensorFlow- Established tech architecture from scratch using MERN stack, React Native, AWS, GCP & Firebase
- Ideated & designed UI/UX for trainer dashboard & cross-platform mobile app (200+ users)
- Owned development, maintenance & deployment for web, Android & iOS apps (10+ trainers, 1k+ MAU)
- Led pose estimation project detecting exercises & counting reps using TensorFlow (20+ exercises)
NAV Fund Administration Group
ReactPython
MySQL
C#
Swiggy
- Analyzed 30+ countries' demographic, regulatory & unit economic data for overseas expansion
- Innovated launch model reducing total cost by Rs60,000/city & managed pilot for 15 cities
Education
BITS Pilani
2016 — 2020Bachelor of Engineering
- GPA: 8.14 / 10
- Coursework: Computer Science, Probability & Statistics, Engineering Optimization, Business Management
- Led 35-member Shell Eco-Marathon team — built India's 1st 100% Ethanol-fueled car
Projects
Where I push AI capabilities to the edge — autonomous agents, real-time systems, and context-heavy architectures.
Kaizen Trader
A case study in context engineering — an autonomous trading engine where Claude Opus analyzes 300 trades + 50 diagnoses per cycle, self-heals bad parameter changes, and evolves strategies through Darwinian selection. The kind of AI system that requires deep understanding of model capabilities, not just API calls.
Momentum, mean reversion, event-driven & structural
Analyzed by Claude Opus + 50 diagnoses per cycle
Circuit breakers on every connection
Railway, Convex, Vercel, GitHub Actions
Wireframr.io
Brand-aware, background-free infographic generator exportable in PNG & SVG. Built with real-time collaboration and AI-powered design suggestions.
Next.jsVercelMarketing Sites
Designed and developed greymetre.io, fieldkonnect.io, and exprosoftech.com — from Figma mockups to production builds.
FigmaGitHub
Where the code lives.
Agents, context pipelines, and AI systems — most of the real work lives in private org repos at @Get-Salesmonk and @EvolvFit.
Travel
30 countries before 30.
Currently 28 years old·1y 10m 28d remaining
A personal mission to explore 30 countries before turning 30 — experiencing cultures, landscapes, and perspectives that shape how I think and build.
Home
🇮🇳IndiaCountries visited
Let's build intelligent systems together.
Gurugram, India
Click to expand
Built with Astro + Three.js