01
Data Science
Working with data through Python, Pandas, NumPy, Matplotlib, and SQL — from cleaning and exploration to useful insights.
NG
AI & Data Science · SATI Vidisha · Pre-Final Year Undergraduate
Nikhil Gumasta.
AIADS Undergrad Data Scientist Aspire Product Manager
I build practical AI tools, data-driven systems, and full-stack applications that solve real problems.
AchievementCounterspell '24 Winner
Currently buildingCareerBoost AI
Continuous grind
DSA · with JAVA
Build the smallest useful thing, then let reality improve it.
01
About
I'm Nikhil Gumasta — Nikk. Pre-Final Year Undergraduate in AI & Data Science at SATI Vidisha. I learn by building practical tools around problems students and teams actually face — data work in Python, useful AI workflows, and full-stack apps.
Right now I’m deep into DSA with Java — learning to break down problems, reason about trade-offs, and write solutions that hold up under pressure. The syntax is only part of it; the real work is learning how to think clearly.
I’m also interested in the space between technical work and product decisions: understanding what people need, choosing a sensible first version, and improving it through real feedback. That keeps me curious beyond the code itself.
02
Skills
A practical toolkit for building intelligent products.
02
Python, Backend & Frontend
Building complete applications with Python, FastAPI, Node.js, React, and JavaScript — from APIs and data flow to clear, usable interfaces.
03
Cloud & DevOps
I use Azure VMs, Ubuntu, Linux CLI, and GitHub to deploy and maintain the applications I build.
04
Java & DSA
Learning DSA in Java right now. Slower than Python but it makes you think more carefully.
03
Tools
The languages and platforms behind my daily workflow.
Languages
Data Science
Frameworks
Cloud & Infra
Dev Tools
Currently Grinding
04
Projects
Selected work, shipped systems, and experiments in progress.
OpenClaw AI agent platform
Self-hosted on Azure VM · Live in production · Telegram + voice + multi-LLM routing
My most involved project. A full AI agent running on an Azure VM — responds on Telegram, makes voice calls via Twilio, handles email. Routes between multiple LLMs depending on the task.
02
Personal AI Assistant
Full-stack · Gemini 1.5 Flash + RAG · Context-aware conversations
Built a personal AI assistant with Gemini 1.5 Flash, FastAPI backend, React frontend, and ChromaDB for context-aware conversations and useful everyday help.
03
LLM fine-tuning dataset
125-sample JSONL · Zero validation errors · Covers reasoning, coding & safety
Wrote a Python script that generates 125 JSONL training samples — reasoning, coding, structured output, safety. Took a while to get zero validation errors but got there.
04 · In progress
BuildingCareerBoost AI
In progress · Resume analysis · Company-specific placement preparation
Building a domain-specific placement assistant that scores resumes, identifies skill gaps, surfaces company-specific interview insights, and creates a week-wise preparation roadmap.
Project details
05
Open Source
Pulled live from GitHub. Some of it is good, some of it is 2am commits.
—
contributions in the last year
@CyberAkhil
Loading contribution graph…