Swapnil Singh
I design and build AI-powered systems, distributed backends, and production-grade products that scale reliably and solve real business problems.
About
- AI Systems Engineer and Full Stack Developer focused on building scalable, reliable, and production-ready AI applications and backend platforms.
- Experienced in designing distributed systems using TypeScript, Python, Kubernetes, event-driven architectures, and vector-based retrieval systems.
- Deeply interested in agentic workflows, multimodal AI, and automation-heavy systems that combine intelligent decision making with clean engineering.
- AI Pipeline Engineer @Fixit, building high-throughput call intelligence systems and automation workflows for real estate brokers.
- Scaled AI backend from early prototype to handling 1000+ calls per hour in production
- Built end-to-end broker automation using LangGraph, WhatsApp, and graph-based RAG
- Technology Lead @Findiy, architecting a distributed AI agents platform for product discovery and outreach in the CPG ecosystem.
- Improved system efficiency by 40 percent through optimized retrieval and agent orchestration
- Designed autonomous multi-agent workflows with clear separation of responsibilities
- Software Engineer Intern @MindFrey, working on AI-assisted learning platforms and recruitment automation.
- Built Q, an AI assistant that guides students through step-by-step problem solving
- Delivered scalable job portal infrastructure using Docker-based deployments
Github Contributions
Stack
Experience
FamApp
Backend Engineering Intern
- Working on backend systems powering recharges, gift cards, and core financial workflows used by 10M+ users.
- Contributing to event-driven services and APIs in a high-scale, production environment.
Fixit
AI Developer Intern
Findiy
Technology Lead
MindFrey
Software Development Engineer Intern
Projects
View AllI built an unofficial MCP server that lets Claude Desktop browse, search, and place orders on Blinkit using plain English. From adding items to the cart to handling checkout and UPI payments, this project turns an AI into a hands-free grocery assistant.
I built an MCP server that lets Claude browse, search, and analyze Reddit in real time. It can explore subreddits, fetch full comment threads, analyze users, and spot trends, all through natural language with zero setup.
I built a Go-based bulk email sender that reads recipients from CSV files and sends emails concurrently using SMTP. It supports HTML templates, retries failed emails with backoff, and is designed to handle large email batches reliably without manual effort.
I built a Twitter-like social platform to understand how social media works at scale. Users can create posts, like, comment, repost, and track their engagement with simple hourly and monthly analytics, all powered by the MERN stack with a clean, responsive UI.