
$ whoami
Guhan Srinivasan
Founder who ships. Registered my company in 2024 and been full-time since. Best part — I'm a business grad, never professionally studied engineering.
I'm building InsdrLink— a vertically integrated technical consulting gig economy for great jobs. We source deals through our own community, render partner and in-house ads inside monetized LLM responses, and plug directly into our partners' SaaS tools by paying for their LLM usage.
5,306 commits in 2026


Altimate Code Hackathon — 2nd Place
I used my RAG pipeline in conjunction with SQL and won 2nd place at the Altimate Code Hackathon.

17 codebases
LLmWiki
A Next.js 15+ application acting as the frontend for your personal AI knowledge base. It uses the app router and is configured with specific AI agent rules (AGENTS.md) to manage structured content generation and display.
Jul 10, 2026
InsdrParse
A multimodal ingestion pipeline for the 'Insdr' ecosystem. It handles OCR for PDFs, transcribes audio files (CAT preparation lectures), and splits large documents into chunks for RAG processing.
Apr 13, 2026
LLm-Wiki-
Public-facing documentation or a mirror of the LLmWiki project, likely intended to showcase the UI or provide user-facing guides for the wiki system.
Apr 10, 2026
insdrdata
A monorepo (Turborepo + pnpm) based on the 'Creatorem' SaaS stack. It is the centralized data management layer for the 'Insdr' project, featuring automated deployment scripts, internal tools for parsing tests, and Supabase integration for database management.
Mar 31, 2026
InsdrLiteRag
A streamlined implementation of LightRAG (Simple and Fast Retrieval-Augmented Generation). It integrates Graph-based RAG with support for various backends like Neo4j, MongoDB, and PostgreSQL to provide context-aware LLM responses.
Mar 23, 2026
UPSC-Data
A search and analysis engine for UPSC Prelims questions. It parses markdown files, uses Google Gemini (2.5-flash) to tag questions by subject/difficulty, embeds them using Azure OpenAI, and stores structured results in PostgreSQL for semantic search.
Mar 18, 2026
PythonBackend
The core production API for 'Insdr', deployed on Google App Engine. It features a hybrid RAG service (rag_service.py), a document registry for managing user uploads in PostgreSQL, and integration with Google Cloud Storage and reranking models.
Mar 2, 2026
CrackMat
An AI-powered search engine and learning platform for GMAT preparation (gmatjs). It integrates Google Gemini for AI-assisted learning and uses 'cheerio' and 'marked' for indexing and rendering educational content.
Feb 20, 2026
brainbackend
A high-performance backend (Node.js/Express) for the 'Brain' project. It handles quick-connect sessions, audio transcriptions, and is fully containerized with deployment scripts for Google Cloud Platform. It uses Prisma for PostgreSQL database interactions.
Jan 2, 2026
brain
The frontend application corresponding to brainbackend. It's a Next.js 15+ app that implements 'NexusWalk' features, integrated with PostHog analytics and designed for complex querying and data visualization.
Jan 1, 2026
NexusQuerying
A dedicated backend for the 'Nexus' engine, focusing on hybrid querying. It supports ingestion and retrieval from Neo4j (Graph) and PostgreSQL (Relational), using Azure OpenAI for embeddings. It includes a multi-user demo for testing isolated knowledge bases.
Dec 31, 2025
literag
The internal development sandbox for the LightRAG framework. It contains the raw source code for testing core RAG features (Neo4j/MongoDB backends, reranking logic, and batch insertion) before they are finalized in the public InsdrLiteRag repository.
Dec 30, 2025
GuhanGPT
A specialized AI chatbot interface and search assistant. It features voice integration (Voice.md), an integration plan for expanded capabilities, and uses Prisma for persistent storage. It acts as a personal or platform-specific GPT implementation.
Oct 9, 2025
GuhanChat
A personal chat application built with Next.js 15+ and Tailwind CSS. It follows your standard app router structure and acts as a playground for custom UI components and chat interactions.
Oct 6, 2025
Odin-KnowledgeGraph
The core knowledge engine for the Odin suite. It manages complex datasets including financial and foreclosure data (Assessor_cases.json), uses Python for building embeddings, and features dedicated query scripts for financial and legal analysis.
Oct 4, 2025
Odin---Agent-JS
A JavaScript/Node.js implementation of the Odin AI agent. It pushes data to 'Graphiti' (a graph database layer), generates person-specific subgraphs, and provides direct Neo4j querying capabilities for real-time agent responses.
Oct 1, 2025
Odin-Graphing
The visualization and graph-building layer for Odin. It features a Python-based graph builder (GPT5Nano.py) and a Next.js dashboard to visualize complex relationships between data entities (like OCR outputs from large documents).
Oct 1, 2025

Altimate Code Hackathon — 2nd Place
I used my RAG pipeline in conjunction with SQL and won 2nd place at the Altimate Code Hackathon.
















