BSc Computer Science @ BITS Pilani
AI/ML + Systems Builder
I like building systems where the interesting part isn't just the model,
but everything around it.
Currently interested in:
→ LLMs & AI Agents
→ Multimodal AI
→ Computer Vision
→ RAG & Retrieval
→ System Design
I enjoy going from:
idea → questionable prototype → "wait, this actually works"
|
A terminal-first development environment inspired by VS Code. Exploring what happens when you take the ideas behind a modern IDE and rethink them for a TUI-native workflow.
|
ML system for detecting/predicting vehicle cut-ins for autonomous driving scenarios. Exploring computer vision and ML for a problem where model predictions have to make sense in a real-world system.
|
|
A Flutter application that keeps track of upcoming competitive programming contests so you don't accidentally discover the contest 17 minutes after it started. Uses the CLIST API for contest data.
|
E-commerce platform for electrical products built during my internship at GrabnGO. Worked across product data pipelines, backend services, search infrastructure and the frontend.
|
┌─────────────────────────────────────────────────────────────┐
│ │
│ 🤖 AI Agents │
│ Multi-LLM systems, tool use, orchestration │
│ │
│ 👁️ Multimodal AI │
│ Vision + language + audio + structured information │
│ │
│ 🔎 Retrieval │
│ RAG, embeddings, semantic search, vector databases │
│ │
│ ⚙️ Systems │
│ Distributed systems, microservices, data pipelines │
│ │
│ 🧠 ML Experiments │
│ Computer vision, RL, evaluation & model experimentation │
│ │
└─────────────────────────────────────────────────────────────┘
Finalist
Finalist
I particularly enjoy hackathons because they're basically an excuse to spend an unreasonable amount of time turning an idea into a working system.
During my internship, I worked on Mahavira Electricals, an e-commerce platform for electrical products.
Some of the things I worked with:
- ⚡ Next.js frontend
- 🔧 Node.js microservices
- 🗃️ DynamoDB product infrastructure
- 🔎 OpenSearch semantic/product search
- ☁️ AWS infrastructure
- 🖼️ Product data & image pipelines
- 🧠 Search designed to understand queries rather than simply match strings
The fun part was getting to work across the entire stack:
Product Data
↓
Data Pipeline
↓
Database
↓
Search / Retrieval
↓
Backend Services
↓
Frontend
↓
Actual Users
LLMs · RAG · AI Agents · Computer Vision · Multimodal AI · Embeddings · Reinforcement Learning
OpenSearch · Vector Databases · Semantic Search
| 🤖 LLMs & AI Agents |
👁️ Computer Vision |
🌐 Multimodal AI |
🔎 RAG & Retrieval |
⚙️ System Design |
I'm particularly interested in the systems around AI — not just making a model produce an answer, but figuring out how to make the whole thing useful, scalable and reliable.
$ whoami
ankit
$ cat interests.txt
AI agents
multimodal models
computer vision
RAG
systems
developer tools
$ cat status.txt
building something
breaking something
learning why it broke
building it again