Engineering
Stack & Radar.
A breakdown of my production engineering stack across Conversational AI, Full-Stack Next.js platforms, Spatial 3D WebGL, and High-Throughput Edge APIs.
AI & LLM Engineering
Architecting production-ready AI agent pipelines, Retrieval-Augmented Generation (RAG), and deterministic tool-calling runtimes with sub-second streaming latency.
Conversational AI Agents
Contextual chatbots, stateful memory, and intent classification (shipped at Kotak MF)
Vercel AI SDK & LangChain
Multi-model orchestration, streaming responses, and LangGraph workflow nodes
Structured Output & Tool Calling
Strict JSON schema enforcement with Zod for deterministic LLM function execution
Vector DBs & RAG
pgvector, Pinecone, and semantic embeddings for high-precision knowledge retrieval
OpenAI & Anthropic APIs
GPT-4o, Claude 3.5 Sonnet integration with custom prompt templates and guardrails
Streaming SSE Runtimes
Server-Sent Events for fluid token-by-token UI rendering without timeout bottlenecks
Frontend & Spatial Web
Crafting high-performance web applications and 60 FPS WebGL experiences with deep attention to render cycles, accessibility, and fluid motion.
Next.js 15 / 14 (App Router)
React Server Components (RSC), Server Actions, Streaming Suspense, and ISR
React 19 & TypeScript
Strict typing, custom hooks, atomic design patterns, and zero type regressions
Babylon.js & WebGL
Interactive 3D indoor wayfinding, coordinate raycasting, and GPU-optimized rendering
Tailwind CSS & Design Tokens
Custom design systems, dark/light themes, and glassmorphic UI architectures
Framer Motion
Buttery spring physics, layout animations, and gesture-driven micro-interactions
Zustand & TanStack Query
Scalable client state management and optimistic cache invalidation strategies
Backend & Edge Runtimes
Engineering resilient microservices, low-latency APIs, and open-source tooling designed for high-concurrency and sub-50ms execution.
Hono.js & Edge Workers
Ultra-lightweight web standards framework; author of @rcmade/hono-docs (10K+ downloads)
Node.js & Express
RESTful APIs, middleware architecture, rate limiting, and robust error boundaries
WebSockets & Realtime
Bidirectional socket channels for live collaboration and streaming analytics
OpenAPI & Swagger Generation
Automated schema extraction and type generation for seamless API-to-client contracts
Python & Fast Data Scripting
Data transformation pipelines and auxiliary ML orchestration services
FinTech Compliance & Auth
Digital KYC workflows, OAuth2, JWT rotation, and SEBI-aligned transaction isolation
Data Systems & Vector Stores
Designing robust schema architectures, relational integrity, in-memory caching layers, and vector indexing for lightning-fast retrieval.
PostgreSQL & pgvector
Relational modeling, indexing strategies, complex joins, and vector similarity search
Supabase & Firebase
Row Level Security (RLS), real-time subscriptions, and scalable managed backends
Redis Cache & Pub/Sub
In-memory session storage, request throttling, and fast distributed caching
Prisma ORM & Drizzle
Type-safe database migrations, declarative models, and connection pooling
MongoDB
Document-oriented storage for flexible schemas and event ingestion feeds
DevOps & Cloud Resilience
Automating CI/CD pipelines, containerized deployments, and edge delivery networks with zero-downtime guarantees.
Docker & Containerization
Multi-stage builds, minimal production images, and container orchestration
AWS Infrastructure
S3, CloudFront CDN distribution, EC2, and serverless compute primitives
Vercel Edge Network
Global edge distribution, custom headers, and automated branch previews
GitHub Actions CI/CD
Automated linting, TypeScript verification, security checks, and production deployments