B.Tech student at IIT Delhi passionate about building scalable backend systems, autonomous AI agents, and crushing algorithmic problems. I turn complex ideas into robust software architectures.
I'm a B.Tech student at IIT Delhi with a deep focus on backend engineering, distributed systems, and applied artificial intelligence. I love building things that scale — from secure REST APIs with role-based access control to autonomous LLM agent frameworks like Hive.
Currently, I am also conducting physics research under Prof. Deepak at IITD, utilizing Computer Vision (OpenCV) to study motility-induced phase separation in active matter. When I'm not architecting systems or running simulations, you can find me sharpening my problem-solving skills in Data Structures & Algorithms and Competitive Programming in C++.
Become a top-tier Software Engineer and push the limits of AI infrastructure.
C++, Python, OpenCV, NLTK, FastAPI
$ whoami
Mukesh Siyol — AI & Backend Systems Dev
$ cat expertise.txt
Python · C++ · OpenCV · AI Agents · NLP
$ cat education.txt
IIT Delhi — B.Tech at IIT Delhi
$ echo $STATUS
✓ Open to internships & full-time roles
$ _
Real-time data fetched from my coding profiles showing consistency and problem-solving skills.
Click any project to see its architecture, tech stack, and what I learned building it.
Production-ready REST API built with FastAPI featuring JWT authentication, role-based access control, analytics endpoints, and full test coverage.
Physics research project under Prof. Deepak analyzing active matter phase separation using OpenCV and Computer Vision algorithms.
Robust data pipeline that scrapes web articles and calculates advanced readability and sentiment metrics using NLTK.
An outcome-driven, evolutionary agent development framework designed to orchestrate complex LLM behaviors.
AI-powered Python tool that parses job descriptions and generates tailored internship application materials.
LLM-assisted PRA COREP regulatory reporting tool with RAG implementation for financial compliance.
Curated collection of optimal DSA problem solutions covering arrays, trees, graphs, and DP algorithms.
A cutting-edge research project conducted under Prof. Deepak in the Department of Physics at IIT Delhi. This research investigates the phenomena of Motility-Induced Phase Separation (MIPS) in active matter.
A robust data pipeline designed to scrape financial and news articles from the web, clean the unstructured data, and calculate highly specific readability and sentiment metrics.
Managing personal finances across multiple streams requires a robust backend that can securely handle transactions, enforce role-based access, and provide actionable analytics without compromising data integrity.
Hive is an outcome-driven agent development framework designed to evolve and adapt. It focuses on abstracting the complexity of orchestrating multiple LLM calls into a cohesive, goal-oriented pipeline.
Applying to hundreds of internships is incredibly time-consuming, especially when tailoring resumes and cover letters to specific job descriptions.
I built an AI-powered tool that automatically parses job descriptions, extracts key skills using NLP, and generates highly tailored internship application materials matching my profile against the job requirements.
Built during an internship, this tool assists in PRA COREP regulatory reporting, which is notoriously complex and strict in the financial sector.
A highly organized repository containing my optimal solutions to essential Data Structures and Algorithms problems, primarily using C++.
Arrays, Linked Lists, Trees, Graphs, Dynamic Programming, Binary Search, Sorting
C++, Python, TypeScript, JavaScript, Java
LLM Agents, RAG Pipelines, NLTK, OpenCV, OpenCV
FastAPI, SQLAlchemy, Pydantic, Node.js, REST APIs
Pandas, BeautifulSoup, HTML, CSS, JavaScript
Git, GitHub, Docker, pytest, CI/CD
A deep dive into how my backend systems are designed — real engineering, real scale.
An evolutionary framework where a swarm of specialized AI agents collaborate to solve complex tasks no single LLM call can handle.
A central controller distributes tasks across Research, Code, and Review agents based on complexity scoring.
ChromaDB vector store enables semantic search across documentation for context-aware code generation.
The Reviewer agent executes code in a sandbox, catches errors, and feeds corrections back to the Coding agent.
An AI-powered tool built during an internship to automate complex PRA regulatory reporting for the banking sector.
Embeds 500+ pages of PRA regulatory PDFs into a vector DB for precise, context-aware LLM queries.
Every generated report passes through strict JSON schema checks before reaching the analyst.
Full traceability of every AI-assisted decision for regulatory compliance and accountability.
My thoughts on AI, Backend Engineering, and Competitive Programming.
Why simple Retrieval-Augmented Generation isn't enough anymore, and how multi-agent frameworks are taking over.
Read Article
Exploring memory management, STL speed, and why Python can't compete in 1-second time limit constraints.
Read Article
How I used computer vision to analyze Motility-Induced Phase Separation during my research at IIT Delhi.
Read Article
I'm actively looking for internship opportunities and full-time roles.
Whether it's a question about AI agents, my Physics research, or just a hello — my inbox is always open!
Fill out the form to send me a direct message, or connect with me via email and LinkedIn below.
MukeshOS v2.0
Type 'help' for a list of commands.
