Hi, I'm Jitin

I build AI agents that can think, see, listen and act.

I'm a BCA graduate, now doing my MCA in AI & ML, who got hooked on one question. What would it take for an AI to actually do things instead of just answering? I'm the creator of Autourgos, an open-source Python framework for LLM agents, and I like building the whole thing end to end, from the reasoning loop to the voice on the other side.

Anime-style illustration of Jitin Kumar Sengar in a dark blazer
Photo of Jitin Kumar Sengar
the real me
Ask me about agents!
17packages on PyPI
27public repos
10achievements and certificates
2podium finishes
1IEEE conference paper
2internships

Chapter 1

A little about me

I'm Jitin, an AI engineer who likes building complete systems rather than single demos. For an agent that means the reasoning loop, the tools it can use, the memory it keeps, and the voice and vision it uses to deal with people.

That's why my work is spread across a few worlds, such as an open-source agent ecosystem (Autourgos), a voice kiosk that answers from an organisation's own documents (Drishti), a desktop assistant you control by talking (ULTRON), and a grading platform that protects students' identity (CampusGrade).

Outside solo projects, I lead the agentic AI work at Team StarkMind, where we build, break and rebuild things at hackathons.

How one of my agents works

  1. It listens and looksVoice (Gemini Live, Whisper) and screenshots as input.
  2. It thinks out loudA ReAct loop of thought, action, observation, repeat.
  3. It does thingsClicks, scrolls, runs tools, and asks me before risky steps.
  4. It remembersVector, episodic and local memory between runs.

Chapter 2

Where I studied

A BCA to learn the fundamentals, and now an MCA to go deep on AI and machine learning.

GLA University campus, Mathura

Completed

Bachelor of Computer Applications (BCA)

GLA University, Mathura

2021 – 2024

K.R. Mangalam University campus, Gurugram

Currently pursuing

Master of Computer Applications (MCA), AI / ML

K.R. Mangalam University, Gurugram

2025 – 2027

Chapter 3

Autourgos, my open-source framework

One small package per capability, pure Python, and it works with any OpenAI-compatible model.

Small, readable, yours to change

Most agent SDKs lock you into one provider or hide the reasoning loop. I wanted the opposite, a tiny core you can read in an afternoon, that runs on OpenAI, Groq, Mistral, DeepSeek or a local Ollama, and stays easy to extend.

  • Works with any OpenAI-compatible LLM, such as OpenAI, Groq, Together, DeepSeek, Ollama and vLLM
  • 11 lifecycle hooks, sync or async, plus approval callbacks for human-in-the-loop
  • Parallel tool calling, and an async twin for every method
  • Pause and resume in-flight runs, with secrets redacted from saved history
pip install autourgos-agent autourgos-openaichat
# Quick start, straight from the README
from autourgos_agent import Agent, tool
from autourgos_openaichat import OpenAIChatModel

# 1. Define a tool
@tool
def calculator(a: float, b: float) -> float:
    """Add two numbers together."""
    return a + b

# 2. Create the agent
agent = Agent(
    llm=OpenAIChatModel(model="gpt-4o"),
    verbose=True,
)
agent.add_tools(calculator)

# 3. Run
result = agent.invoke("What is 123 + 456?")
print(result)
# 579

Chapter 4

Things I've built

The bigger stories first, then the smaller experiments.

Chapter 5

Future projects

What I'm working towards next.

Coming soon

RAG Studio

A platform that crawls online sources and documents, then cleans, checks and validates the data before it reaches a production RAG system. The goal is to turn raw content into reliable, RAG-ready knowledge.

  • Crawl
  • Clean
  • Check
  • RAG

Chapter 6

My achievements

Two podium finishes at our university, a top 5 finish at a national hackathon, four events I took part in, two internships, and a paper at an IEEE conference.

Awards 3

Third position

INTELLITHON'26

20 – 21 April 2026 · K.R. Mangalam University, Gurugram

We presented VoxAssist, an AI voice assistant focused on natural interaction, automation and productivity, and took third place at the university hackathon.

Team Lead: Sonia
Members: Jitin Kumar Sengar and Khushi Nagpal

Third position

Best Project Award

24 April 2026 · K.R. Mangalam University, Gurugram

We presented Vox Agent, an agentic AI lab for autonomous decision-making, task execution and user interaction, and placed third for innovation, creativity and practical impact.

Team Lead: Jitin Kumar Sengar
Members: Sonia, Sadgyan Singh, Anshika, Ashish Raj and Shashi Kant

Top 5 · Semi-finalist

Build with Bharat 2.0

13 – 14 September 2026 · NIT Delhi

We qualified as semi-finalists in Build with Bharat 2.0, a national level hackathon organized by Codeverse at NIT Delhi, and our team, Team StarkMind, finished in the top 5.

Team Lead: Sonia
Members: Jitin Kumar Sengar, Ishika Batra and Atharv Kashiv
Certificate: Certificate of Appreciation

Participation 4

Participation

Deep Data Science Hackathon 2.0

8 October 2025 · K.R. Mangalam University, Gurugram

I took part in the Deep Data Science Hackathon 2.0, organized by the Center of Excellence AI at K.R. Mangalam University.

Certificate: Certificate of Participation

Participation

Code Canvas

11 November 2025 · K.R. Mangalam University

I took part in Code Canvas, a Visual AI Tool Design event organized by K.R. Mangalam University.

Certificate: Certificate of Participation

Participation

Manthan AI 2025

26 – 28 November 2025 · K.R. Mangalam University

I took part in Manthan AI 2025, the 1st AI Summit, organized by the Center of Excellence AI and the School of Engineering and Technology at K.R. Mangalam University. It had keynote talks, hands-on AI workshops, alumni sessions and a competitive hackathon. Our team proposed Bhramsatra, an agents module with a tool call agent, a multi tool call agent and a ReAct agent.

Team Lead: Jitin Kumar Sengar
Members: Sonia and Aman
Certificate: Certificate of Participation

Participation

Hack KRMU 5.0

18 – 20 February 2026 · K.R. Mangalam University, Gurugram

I took part in Hack KRMU 5.0, a three day hackathon organized by K.R. Mangalam University.

Team Lead: Jitin Kumar Sengar
Members: Sonia, Yashwant Giri and Vishakha
Certificate: Certificate of Participation

Internship 2

Internship · Kyoto, Japan

Sakura Immersion Program at Kyoto University

21 June – 2 July 2026 · Kyoto University, Japan

I took part in this special program on Smart Sustainability, AI & ML for Global Impact, run at Kyoto University by Kyoto University Original Co., Ltd., while studying at K.R. Mangalam University. My university recognises this program as my internship.

Certificate: Certificate of Attendance

Internship

AI/ML Engineering and Problem Solving Bootcamp

About 2 months · Completed 24 July 2026 · K.R. Mangalam University

I attended this bootcamp, run by Programming Pathshala at K.R. Mangalam University, and my university counts it as my internship. It covered Python, data analysis and visualization, machine learning, deep learning, generative AI, LLMs, AI application development, SQL, data structures and algorithms, and computer science fundamentals.

Certificate: Certificate of Completion

Research paper 1

IEEE conference paper

Singleton Connection Pooling for Low-Latency Gemini Orchestration, A Comparative Benchmark

24 – 26 September 2026 · K.R. Mangalam University, Gurugram

I presented this paper at MRIE-2026, the 3rd International Conference on Multidisciplinary Research and Innovations in Engineering, organized by the School of Engineering and Technology at K.R. Mangalam University. It is a comparative benchmark of connection pooling for low-latency Gemini orchestration.

Author and Presenter: Jitin Kumar Sengar
Co-author: Sonia
Certificate: Certificate of Presentation

Chapter 7

What's in my toolbox

Each group comes with the places I actually used it.

Agents

Used in Autourgos, ULTRON, Bhramsatra at Manthan AI

  • ReAct agents
  • Tool calling
  • Multi tool calling
  • Middleware hooks
  • Episodic memory
  • RAG
  • Human-in-the-loop
  • Computer-use

AI & machine learning

Used in my MCA, the AI/ML bootcamp, Deep Data Science Hackathon, Kyoto program

  • Python
  • Machine learning
  • Deep learning
  • Generative AI
  • LLMs
  • Data analysis & visualization

LLMs & realtime

Used in VoxAssist, Examina, my IEEE paper

  • OpenAI APIs
  • Gemini Live
  • Gemini Vision
  • Faster-Whisper
  • WebSockets
  • Structured outputs
  • Latency benchmarking
  • Connection pooling

Backend

Used in CampusGrade

  • Python (async)
  • FastAPI
  • PostgreSQL
  • Docker
  • S3 storage
  • Caddy HTTPS

Frontend & mobile

Used in Drishti kiosk, this portfolio

  • TypeScript
  • React
  • Next.js
  • Tailwind CSS
  • Kotlin
  • Jetpack Compose
  • HTML & CSS
  • Vanilla JavaScript

Programming & computer science

Used in the AI/ML bootcamp

  • C
  • C++
  • Java (core)
  • Python
  • Data structures & algorithms
  • SQL
  • Problem solving
  • CS fundamentals

Tools I work with

AI tools

  • Claude Code
  • Codex
  • Antigravity
  • Ollama

Desktop apps

  • Qt
  • Tkinter
  • PyInstaller

Code and hosting

  • Git
  • GitHub
  • GitHub Pages
  • CI/CD

Final chapter

Let's talk

I'm open to agentic AI roles, collaborations and interesting research. Tell me what you're working on.

This opens your email app with the message filled in.