FULL-STACK & AI SYSTEMS DEVELOPER

I BUILD SECURESYSTEMS ANDINTELLIGENT TOOLS.

Explore work

Software, engineered
with clarity.

Full-stack developer working across Java, Spring Boot, React, TypeScript, and Python. I build secure APIs, AI-assisted applications, and developer tools that turn complex workflows into dependable products.

OPEN SOURCESubmitty contributor
JAVA / SPRINGREACT / TYPESCRIPTMCP / AGENTSDOCKER / REST APIs
Scroll to explore

ABOUT ME (&)

FROM VISUAL DESIGNTO ENGINEERING SYSTEMS.

Visual communication taught me clarity. Full-stack work turned that instinct into systems; open source and AI tooling keep sharpening it.

  1. 2024

    Visual foundations

    Learning to communicate through design

    Worked as a freelance graphic designer, developing practical experience in visual hierarchy, creative interpretation, client communication, feedback, and independent delivery.

    FORM → HIERARCHY → CLARITY
  2. 2025

    Full-stack & backend

    Moving from interfaces into complete systems

    Built React products and layered Java services with authentication, persistence, validation, payments, API contracts, and clear domain boundaries.

    CONTROLLER / SERVICE / REPOSITORY
  3. 2025

    Machine learning foundations

    Understanding the foundations behind intelligent products

    Completed the Machine Learning Specialization covering supervised learning, advanced learning algorithms, unsupervised learning, recommender systems, and reinforcement learning.

    DATA → MODEL → EVALUATION
  4. 2025–PRESENT

    Open source

    Learning inside a large collaborative codebase

    Contributing to Submitty through issue investigation, interface fixes, pull-request testing, review discussions, and maintainer collaboration.

    ISSUE → ROOT CAUSE → REVIEW
  5. 2026

    AI developer tools

    Building reliable software around models

    Building AI-assisted workflows, MCP servers, developer tooling, deterministic safety controls, and systems where probabilistic models do not replace verification.

    MODEL ≠ AUTHORITY

SELECTED WORK

SYSTEMS BUILTTO BE UNDERSTOOD.

AI-assisted products, developer tools, backend services, and full-stack applications built around clear architecture, safety, and dependable behavior.

01AI-assisted healthcare prototype

SignalCare

AI-assisted triage with deterministic safety guardrails

  • GEN AI
  • SAFETY
  • REACT
02AI-assisted developer tooling

PatchPilot X

From engineering alert to evidence-backed patch

  • AI AGENTS
  • DEV TOOLS
  • FASTAPI
03Model Context Protocol server

Zulip MCP

A safe, topic-aware MCP server for team communication

  • MCP
  • TYPESCRIPT
  • AI TOOLS
05Full-stack product

TOMATO

A food-ordering experience from discovery to delivery

  • REACT
  • NODE.JS
  • MONGODB

Scroll to move through selected work →

CAPABILITIES

WHAT I BRINGTO THE SYSTEM.

Secure backends, useful interfaces, and controlled AI workflows — tested at the seams.

Layered services, secure APIs, clear domain boundaries, predictable validation, persistence, and maintainable business logic.

  • Java
  • Spring Boot
  • Spring MVC
  • Spring Security
  • MongoDB
  • SQL

EXPERIENCE

TWO DISCIPLINES.ONE ENGINEERING PRACTICE.

01

December 2025–Present

Open-Source Contributor

Submitty

Contributing to a large open-source education platform through issue investigation, frontend fixes, discussion-forum improvements, grading-interface edge-case handling, pull-request testing, code review, and maintainer collaboration.

  • Bug investigation
  • Root-cause isolation
  • Targeted changes
  • Peer-PR testing
  • Code review
  • Collaborative Git workflows
02

November 2024–September 2025

Freelance Graphic Designer

Upwork

Delivered visual-design work while managing client communication, creative interpretation, revisions, and independent delivery—the visual foundation that still informs how I structure software interfaces.

  • Visual hierarchy
  • Creative interpretation
  • Client communication
  • Revision cycles
  • Independent delivery

OPEN-SOURCE PROOF

CONTRIBUTING BEYONDMY OWN CODEBASE.

A five-step method for entering an unfamiliar codebase without guessing.

01 / 05
01

Investigate

Reproduce the issue and map the affected interface or backend path.

02

Isolate

Reduce the failure to its root cause and identify the smallest responsible boundary.

03

Change

Implement a targeted fix with attention to nearby behavior and edge cases.

04

Verify

Test the change and peer pull requests in the real project workflow.

05

Review

Discuss tradeoffs with maintainers and iterate on feedback.

Verified contribution URLs remain intentionally hidden until supplied; no issue, pull-request, or impact counts are invented.

CURRENT DIRECTION

BUILDING TOWARDPRODUCTION AI.

A four-term path from model foundations to dependable AI systems.

01Starting soon

Computer Vision

End-to-end visual pipelines, deployment, evaluation, monitoring, and robust model behavior.

02Upcoming

NLP & Transformers

Pretrained models, transformer systems, fine-tuning, evaluation, and production integration.

03Upcoming

Multi-Agent Orchestration

Agent roles, tools, shared state, coordination, failure handling, and evaluation.

04Upcoming

ML Systems & LLMOps

Reproducibility, versioning, monitoring, feature stores, deployment, and model CI/CD.

EDUCATION & CREDENTIALS

FOUNDATIONS,
DOCUMENTED.

EDUCATION / IN PROGRESSEXPECTED 2028

Currently pursuing

Bachelors of Science

Computer Science

Birla Institute of Technology and Science, Pilani

CERTIFICATE / COMPLETEDOCT / 2025

3-course specialization

Machine Learning Specialization

DeepLearning.AI and Stanford Online · Coursera

  • Supervised Machine Learning: Regression and Classification
  • Advanced Learning Algorithms
  • Unsupervised Learning, Recommenders, Reinforcement Learning

FAQ

A FEW THINGSYOU MAY WANT TO KNOW.

Backend, full-stack, AI engineering, developer-tooling, open-source, internship, and selected freelance opportunities.

No. Each case study identifies whether a project is deployed, open source, production-style, experimental, or educational. Limitations are documented instead of hidden.

AI is used for language understanding, drafting, extraction, or automation. Critical decisions such as safety escalation or test verification remain deterministic wherever reliability requires it.

Yes. Case studies link to public repositories and explain important architecture and entry points.

START A CONVERSATION

BUILDING SOMETHINGTECHNICALLYCHALLENGING?

I’m interested in backend systems, AI developer tools, open-source software, and products where reliability matters.