CS undergraduate at SRM (CGPA 9.55/10) & founder of Gradely — building agentic systems that understand, debug, and refactor real production software with local LLMs and control-flow reasoning.
Great developer AI doesn’t just autocomplete boilerplate —
it reasons about control flow,
patches bugs, and proves itself.
Deep code understanding, AST program analysis, automated debugging, and multi-step refactoring across real codebases. Orchestrates local LLMs using LangGraph state machines and calls tools through the Model Context Protocol (MCP).
End-to-end production RAG pipeline covering web scraping, semantic chunking, metadata filtering, and grounded synthesis with BGE embeddings. Benchmarked ChromaDB, Pinecone, and FAISS on relevance, faithfulness, and latency.
Autonomous SRE platform engineered for Kubernetes clusters. Features a deterministic root-cause engine with local LLM diagnostics that chooses and executes remediations such as automated pod restarts, rollbacks, and horizontal scaling.
MSME registered company with production LLM pipelines automating client communication, milestone tracking, and automated payout management. Scaled full-stack event-driven services to 100+ clients with >₹26,000 revenue.
A tri-modal recommender system architected with Vision Transformers (ViT-B/16), EfficientNet-B4, and KMeans clustering. Features a dual-stream pipeline that analyses facial morphology, achieves 91.2% skin tone classification accuracy, and fuses real-time ambient weather data to generate grounded, personalised outfit recommendations.
BlameBot runs locally without leaking proprietary codebase contexts. Using LangGraph state machines and the Model Context Protocol (MCP), it inspects Abstract Syntax Trees, runs tests, and applies surgical patches with verifiable guarantees.
parse_config()Code generation, AST & program analysis, repository reasoning, agentic debugging, LangGraph, Model Context Protocol (MCP), local Ollama inference.
→PyTorch, TensorFlow, Hugging Face Transformers, Vision Transformers (ViT), EfficientNet, Computer Vision, OpenCV, NLP pipelines.
→Retrieval-Augmented Generation, LlamaIndex, ChromaDB, FAISS, Pinecone, BGE embeddings, metadata filtering, chunking optimization.
→Python, C++, Rust, TypeScript, Go, JavaScript, SQL. Data Structures & Algorithms, OOP, DBMS, OS internals, Distributed Systems.
→FastAPI, Kubernetes, Prometheus, Docker, Linux, CI/CD pipelines, AWS (EC2, S3, Lambda), Oracle Cloud Infrastructure (OCI).
→React.js, Node.js, Express, MongoDB, PostgreSQL, event-driven architectures, REST APIs, high-throughput data processing pipelines.
→Sep 2025 — Present | Gradely (MSME Registered). Engineered LLM pipelines automating client communication, milestone verification and payouts for 100+ clients.
Jul 2025 — Sep 2025 | Saday Sadev. Architected automated SQL & Python ingestion pipelines for 500+ stakeholders, reducing manual wrangling by 8+ hours/week.
Jan 2024 — Present | GitHub Community. 20+ merged pull requests across 3 repositories: bug fixes, performance profiling, and control-flow reasoning.
B.Tech Computer Science Engineering @ SRM Institute of Science & Technology. CGPA: 9.55/10. Smart India Hackathon Semi-finalist 2025.
Oracle Certified Foundations Associate validating comprehensive understanding of Agentic AI, autonomous multi-agent systems, and generative pipelines.
Semi-finalist in Smart India Hackathon 2025, architecting technology-driven solutions to solve high-impact national infrastructure challenges.
Selected participant in the McKinsey Forward Program for digital leadership, along with Big Data & Data Science Bootcamp certification by C-DAC.
Open to internships, full-time engineering roles, and research collaborations in AI/ML, autonomous agents, and software engineering. Reach me directly at rishabhsingh301220@gmail.com or send a message below.
New Delhi, India · Responses typically within 24 hours.