Dark Mode

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

Sign up
Appearance settings
dipakkr
Follow
Working remotely

Deepak Kumar dipakkr

Working remotely
Software Engineer | Building micro saas products

Organizations

Block or report dipakkr

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user's behavior. Learn more about reporting abuse.

Report abuse
dipakkr/README.md

Hey, I'm Deepak

Senior AI Engineer * LLM Training & Evaluation * RAG * Agentic Systems * Distributed Systems at Scale

Bengaluru, India * dipakkr.co@gmail.com * tooljunction.io

Trained LLaMA 4 with researchers at Meta, OpenAI & Google. Scaled real-time infrastructure to 100K+ concurrent users. Maintaining open-source repositories with 24K+ GitHub stars.


What I've Shipped

System Scale Result
LLaMA 4 Training & Evaluation (Turing / Meta / OpenAI) 1,000+ training examples +23% code gen accuracy * -15% hallucination rate
Wrike AI Portal -- Vertex AI RAG System 10K+ daily queries -40% knowledge discovery time
Wrike Herald MCP Server Org-wide adoption -90% manual dev work during migrations
Wrike AI Prototype Generator Design - Dev pipeline -60% turnaround time
Airmeet Real-Time Event Infrastructure 100K+ concurrent users <100ms latency * Kafka * AWS Lambda
Airmeet GenAI Features (GPT-4, Whisper API) 500+ hours/month saved Automated summarisation + transcription

Open Source AI Portfolio

Project Description
AI Code Reviewer with Memory AI-powered code review system with team memory -- recalls past PR decisions, coding conventions, and rejected patterns. Built on MCP + vector DB
LLM Evaluation Framework Open-source framework to test LLMs (GPT-4, Claude, Gemini, LLaMA) against custom test suites -- measuring hallucination rate, reasoning quality, and code correctness with a built-in dashboard
Prompt Diffing Tool Developer tool to visually diff two prompt versions and see how output behavior changed across test cases, with statistical significance -- like git diff for prompts

Other Projects

Project Description Stars
A to Z Resources for Developers One of India's most-starred open-source dev learning repositories 22K+
ClaudeAI.directory Curated directory of Claude AI tools and resources 2K+
Tooljunction AI tools marketplace with GPT-4 & Gemini powered recommendations --
PostZaper.com AI-powered bookmarking and content generation --
AILLMJobs.com Curated job board for AI, LLM, and machine learning roles --
DevKitOS.com Open-source developer toolkit and resource hub --

Core Stack

AI & LLMs

Languages & Backend

Infrastructure & Cloud


GitHub Stats


Pinned Loading

  1. A-to-Z-Resources-for-Students A-to-Z-Resources-for-Students Public

    Curated list of resources for college students

    21.4k 4.8k

  2. node-auth node-auth Public

    Implementing Authentication in NodeJS With Express and JWT - CodeLab #1

    JavaScript 88 43

  3. 3d-cnn-action-recognition 3d-cnn-action-recognition Public

    Implementation of Action Recognition using 3D Convnet on UCF-101 dataset.

    Python 76 32