Skip to content
View pradhyumna186's full-sized avatar

Block or report pradhyumna186

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
pradhyumna186/README.md

Hi, I'm Pradhyumna πŸ‘‹

Software Engineer β€” Backend & AI Systems

I build backend services and AI-powered developer tools β€” Spring Boot / FastAPI systems on one side, agentic LLM workflows on the Anthropic Claude API on the other. M.S. Computer Science, University of Florida (GPA 3.92), graduated Dec 2025. Open to full-time software engineering roles.

What I focus on

  • Backend & systems β€” Java / Kotlin / Spring Boot, Python / FastAPI, REST & GraphQL, microservices, performance profiling and optimization
  • Agentic AI β€” Anthropic Claude API, prompt & context engineering, LangGraph multi-agent systems, RAG, MCP, LLM-as-judge evaluation
  • LLM internals β€” implemented GPT-2 from scratch (multi-head attention, transformer blocks, training loop) to understand what's actually under the hood

πŸ›  Tech Stack

Languages

Java Kotlin Python TypeScript C++ SQL

Backend & AI

Spring Boot FastAPI GraphQL React Anthropic Claude LangGraph

Data & Infra

PostgreSQL MySQL Redis MongoDB AWS GCP Docker Kubernetes GitHub Actions


πŸš€ Featured Projects

Python Β· FastAPI Β· Anthropic Claude API Β· Strawberry GraphQL Β· Redis Β· GitHub Actions

  • Uses the Anthropic Claude API to analyze pull-request diffs across diverse, unfamiliar codebases and generate structured, context-aware review feedback.
  • Engineered prompts and structured-output schemas with an LLM-as-judge evaluation pipeline that catches hallucinations before anything is posted.
  • Deployed on Cloud Run with multi-tier Redis caching; shipped with automated tests and CI/CD.

Python Β· LangGraph Β· FastAPI Β· PostgreSQL Β· GPT-4o

  • Multi-agent LangGraph system with dynamic tool/function calling and self-reflection, orchestrating LLM reasoning over live data with conditional branching and state management.
  • Optimized prompts and context windows to cut per-request LLM cost ~35% while preserving quality; added an LLM-as-judge framework for citation accuracy.

πŸ’Ό Experience

Open Source Software Developer (Android) Β· Greenstand Β· Jan 2026 – Present Refactored a localization engine in a large open-source codebase to fix a critical UI desync bug and resolved an intermittent production crash via dependency-injection scope management (Koin).

Software Engineer Intern (Performance & Backend) Β· Superstars, Inc. Β· Aug – Dec 2025 Profiled and optimized production systems (JProfiler) and built fault-tolerant Java / Spring Boot microservices with automated testing and CI/CD.

Machine Learning Intern Β· Jio Platforms Ltd. Β· Jul – Sep 2023 Built distributed PySpark / XGBoost data pipelines processing 10M+ daily records.


πŸŽ“ Education

  • M.S. Computer Science β€” University of Florida Β· 2024 – 2025 Β· GPA 3.92
  • B.Sc. Computer Science β€” Bennett University, India Β· 2020 – 2024 Β· GPA 3.83

πŸ“« Connect

LinkedIn GitHub Email

Top Languages

Popular repositories Loading

  1. mailTide mailTide Public

    Python 1

  2. MARKS-MANAGEMENT-SYSYTEM MARKS-MANAGEMENT-SYSYTEM Public

    Java

  3. Vehicular-Pollution-Analysis-using-ML Vehicular-Pollution-Analysis-using-ML Public

    Forked from surya-teja-222/Vehicular-Pollution-Analysis-using-ML

    Pollution Monitor

    Java

  4. pradhyumna186 pradhyumna186 Public

  5. product-preview-card-component-main product-preview-card-component-main Public

    CSS