Software Engineer / AI & ML

Building useful systems with intelligence.

Engineer focused on full-stack products, distributed systems, and applied AI. I like turning ambiguous problems into production-ready software.

View selected workGet in touch ↗Download Resume ↓
New Delhi, India
BTech CSE (AI & ML)
K.R. Mangalam University
2027
Portfolio / Vol. 01
01 / About

A builder at the intersection of software and AI.

I work across frontend, backend, data, and modern AI systems—moving from product idea to architecture, implementation, and delivery. My recent work spans collaborative editors, event-driven platforms, computer vision pipelines, and production APIs.

Currently studying Computer Science & Engineering with a specialization in Artificial Intelligence & Machine Learning.
02 / Capabilities

The technical toolkit.

Core CS

Data Structures & Algorithms, OOP, System Design, Distributed Systems.

Languages

Java, Python, JavaScript, TypeScript, SQL.

Frontend

React.js, Next.js, HTML5, CSS3, Tailwind CSS, State Management.

Backend

Node.js, Spring Boot, FastAPI, REST APIs, GraphQL, Microservices, API Design, Kafka.

Data

PostgreSQL, MySQL, MongoDB, Redis, Data Modeling, Database Design, Caching.

AI / LLM

Generative AI, AI Agents, RAG, OpenAI, LangGraph, Tool Calling.

Platforms

Git, GitHub, Jira, Docker, Kubernetes, CI/CD, Agile/Scrum.

03 / Experience

Building in teams.

SDE Intern
Bluestock
  • Spearheaded a team of 9 interns to build a production-ready IPO web application and RESTful API using React, Node.js, Kafka, and SQL, enabling tracking of 50+ IPOs with dynamic price bands, listing gains, and reports.
  • Optimized backend performance by 40% through efficient SQL queries and Redis caching, improving scalability, response times, and data accessibility.
  • Led agile workflows using Git, GitHub, and Jira, achieving 100% on-time delivery and a 30% boost in development velocity.
May 2025
— Jun 2025
Data Scientist Intern
Zidio Development
  • Engineered a vision-language pipeline integrating BLIP for image captioning and UNet for semantic segmentation of 1,000+ images.
  • Enhanced multi-modal model accuracy by 25% using vision transformers and encoder-decoder architectures.
Jun 2025
— Aug 2025
04 / Selected Work

Systems worth shipping.

01 / Collaborative Software

AI-Powered Collaborative Document Editor

A Google Docs–style collaborative editor with real-time editing, AI-powered document generation and summarization, role-based access control, organization-level document management, commenting, notifications, and export.

RealtimeAIRBACDocs
02 / DevOps Intelligence

CICD-AI: Event-Driven Failure Intelligence Platform

An event-driven microservices platform orchestrating Spring Boot, Kafka, Docker, PostgreSQL, and React. Includes automated build workers, webhook ingestion, and LLM-powered root-cause analysis for real-time CI/CD pipeline failure intelligence.

KafkaMicroservicesLLMCI/CD
03 / Predictive Maintenance

Lakehouse-Based Predictive Maintenance

AssetPulse, an end-to-end predictive maintenance platform using Python, SQL, DuckDB, dbt, Airflow, and Docker to process NASA C-MAPSS sensor data through a Bronze–Silver–Gold lakehouse architecture, with automated ETL pipelines, data validation, feature engineering, and interactive analytics dashboards for industrial equipment monitoring and predictive insights.

LakehouseAirflowDuckDBPredictive AI
04 / On-Device Voice AI

VoxRoute: On-Device Real-Time Voice AI Agent

An on-device voice pipeline orchestrating Silero VAD, Faster-Whisper STT, ChromaDB vector search, and Ollama LLM. Implements state-machine transitions, deterministic tool execution, and groundedness guardrails for real-time travel support.

Voice AIWhisperRAGOn-Device
05 / Recognition

Proof in the numbers.

103TCS CodeVita S13 — global rank
Top 100India — College Youth Ideathon
4×Hackathon winner
1600+LeetCode rating
06 / Education

Learning with intent.

2027
BTech CSE (AI & ML)
K.R. Mangalam University
2023
High School
Kamal Model Sr. Sec. School
The portfolio intentionally keeps the résumé’s terminology and claims intact, while turning them into a more editorial reading experience.
07 / Contact

Let’s build something useful.

Open to engineering, AI/ML, and product-building opportunities where strong systems thinking meets real-world delivery.