Piyush MakadFull Stack AI Engineer · India / Remote
Production LLM products · React to backend · Distributed systems
At ChatDaddy, I built customer-facing LLM workflows, a 120-tool MCP platform, and AI-assisted support systems. I own the React experience, model and tool integration, backend architecture, and production reliability.
Career & selected work
Production LLM features, tool orchestration, support automation, and the high-volume infrastructure behind them.
AI tools · ChatDaddy
Connected AI clients to messaging, CRM, automation, and commerce with typed tool schemas, team-scoped access, and approval-gated writes.
Full-stack AI · ChatDaddy
Built a production React and OpenAI Responses experience that converts natural language into structured automation workflows.
AI analytics · ChatDaddy
Owned analytics comparisons, LLM prompts and fallback, billing, email delivery, and the React experience for a paid report product.
AI-assisted support · ChatDaddy
Originated an investigation system adopted into a production support tool that correlates Kibana, frontend, and backend evidence.
Data architecture · ChatDaddy
Separated active and historical message access with MongoDB hot and cold partitions and Elasticsearch-backed search.
API performance · ChatDaddy
Traced repeated contact operations in production and optimized the lookup path, making contact access 5× faster.
Production reliability · ChatDaddy
Investigated a high-failure retry path and improved message handling, taking decryption failures from 65% to under 1%.
Real-time systems · CineStox
Built a Node.js order-book service using Kafka, Socket.IO, PostgreSQL, and Redis for order processing and live updates.
Experience
ChatDaddy
Shipped a 120-tool MCP platform, AI prompt-to-flow generation, paid Daily AI Reports for 3,000+ teams, and AI-powered WhatsApp Business insights for 2,000+ teams. Built React interfaces and model-backed services while optimizing high-volume messaging, data access, and incident response.
CineStox
Built a Kafka and Socket.IO order-book MVP, introduced Redis caching that reduced database load by 40%, and shipped payment and notification services.
SKIT Jaipur · 8.90 CGPA
Built a foundation in software engineering, algorithms, databases, and system design.
Independent engineering
AI products with working interfaces, plus PulseFlow: a completed Go system for durable event ingestion and notification delivery.
Built a multi-tenant Go event and notification system with authenticated APIs, a PostgreSQL outbox, Kafka, idempotency, bounded workers, and durable retries.
View repositoryRepository summaries, semantic code search and meeting transcription in one developer workspace.
Open productVoice interview practice with role-aware conversations and structured feedback.
Open productBuilding in public
How I can help
Model integration, React interfaces, backend services, and reliable operation after launch.
Building React experiences around LLM workflows, analytics, and automation, with backend integration, billing, and delivery.
Connecting AI clients to real product actions with typed MCP tools, team-scoped access, and approval for sensitive writes.
Designing event flows and data access for high-volume products, then improving performance, reliability, and incident response.
Technical toolkit
I build across React, TypeScript, LLM workflows, and MCP, with deep experience in Node.js, event-driven services, databases, and production debugging.
Let's build something
I'm open to Full Stack AI Engineer roles where I can ship useful AI products and own the backend systems that make them dependable.