Software Engineer
Open to full-time SWE roles

Hi, I'm Karanbir Singh

Karanbir Singh Pelia — Software Engineer and MS Computer Science graduate of Stony Brook University
Karanbir Singh Pelia
Software Engineer
0 Projects
0 Roles
0 Publication
2026 MS CS Grad

Software Engineer building reliable backend and distributed systems, RAG/LLM-powered tools, and real-time ML applications. MS in Computer Science from Stony Brook University, graduated May 2026.

About

A bit about myself

I'm a Software Engineer with an MS in Computer Science from Stony Brook University, passionate about building reliable backend systems, distributed platforms, and automation tools. I enjoy turning complex problems into clean, well-tested solutions.

My experience spans building REST APIs with Spring Boot and FastAPI, crafting RAG/LLM-powered applications, and developing ML models for real-world use cases. I care deeply about code quality, observability, fault tolerance, and iterative development.

I'm currently open to full-time Software Engineering roles — feel free to reach out or download my resume.

Skills & Stack
Languages
Python Java C++ JavaScript SQL
Backend
Spring Boot FastAPI Flask REST APIs Pydantic Kafka gRPC RAFT PostgreSQL Redis MongoDB MySQL
Frontend
React Tailwind CSS HTML
Data Engineering
Confluent Schema Registry Avro Parquet PyArrow
Testing & DevOps
JUnit 5 Mockito Pytest Docker Kubernetes Jenkins CI/CD GitHub Postman
Observability
Prometheus Grafana Micrometer
AI / ML
Ollama / LLMs RAG ChromaDB Claude Code Copilot Cursor Codex TensorFlow OpenCV NumPy Prompt Engineering
Education

Academic background

Stony Brook University
Master of Science — Computer Science
Stony Brook, NY Aug 2024 – May 2026 3.62 GPA
Coursework
Data Science HCI Computer Graphics Advanced Algorithms Distributed Systems Data Management
Savitribai Phule Pune University
Bachelor of Engineering — Computer Engineering
Pune, MH, India Sep 2020 – Jun 2024 3.64 GPA
Coursework
Data Structures Algorithms OOP Operating Systems Computer Networks Database Systems Machine Learning
Experience

Professional journey

Software Engineer Intern
Infrrd Inc.
Internship May 2025 – Jan 2026 San Jose, CA
Built an internal FastAPI evaluation platform automating accuracy benchmarking across document processing pipelines, standardizing QA workflows and cutting manual effort roughly in half.
Strengthened an LLM-based document verification framework, resolving misclassification issues and lifting extraction accuracy on key document fields by over 20 percentage points.
Shipped feature-flagged production releases with output validation and fallback handling, enabling zero-downtime rollouts across Agile sprints.
Built a stateless async replay endpoint for classification testing, removing the need to re-run the full pipeline and speeding up debugging and QA turnaround.
Python FastAPI Flask REST APIs MongoDB Jenkins Grafana JavaScript Jira Confluence Postman Copilot Claude Code
Graduate Research Assistant
Stony Brook University
Research Jan 2025 – Dec 2025 Stony Brook, NY
Exploring visual storytelling with Prof. Klaus Mueller to enhance infographic communication through iterative feedback loops.
Evaluating LLMs' alignment with human cognition in interpreting data visualizations and complex information displays.
LLMs Data Visualization HCI Python
Software Developer Trainee
Campus Credentials
Trainee Jan 2023 – Feb 2023 Pune, MH
Strengthened backend development fundamentals including OOP concepts, DSA, RESTful APIs, web scraping, and multithreading.
Built a Python CLI application integrated with MySQL to execute CRUD operations for managing structured student data.
Python MySQL REST APIs OOP DSA
Projects

Featured work

A collection of things I've built — from distributed backend systems to automation pipelines and real-time ML applications.

Entitle

Privacy-first RAG agent that turns a free-form description of someone's household into a structured profile and matches it against 15 federal and 14 state safety-net programs, surfacing $2,500+ in monthly benefits. Runs fully offline via Ollama, with a document reader that explains photographed benefit notices and denial letters in plain language.

Python FastAPI React Tailwind CSS Gemma 4 ChromaDB Ollama
MediGrid

Fault-tolerant medical resource availability platform built with Spring Boot, Kafka, gRPC, RAFT, PostgreSQL, Redis, Docker, and Kubernetes. Implements RAFT consensus from scratch to maintain write availability under node failures, validates streamed updates against Avro schemas, and powers OLAP analytics through a Parquet-exportable history store.

Java Spring Boot Kafka Avro gRPC RAFT PostgreSQL Redis Docker Kubernetes Prometheus
AgentFlow

Turns plain English into executable API workflows. Local Llama 3 reasons over tasks, outputs validated JSON payloads, and calls external REST APIs end-to-end with zero cloud dependency.

Python Ollama Llama 3 FastAPI Pytest
TrackDesk

Production-grade ticket management API built with Java and Spring Boot. Fully observable via Micrometer, Prometheus, and Grafana, with 80%+ test coverage using JUnit 5 and Mockito, plus Swagger documentation.

Java Spring Boot Prometheus Grafana JUnit 5 Swagger
Site to Slides

Paste a URL, get a ready-to-present slideshow. Reverse-engineered undocumented API endpoints via browser network inspection to automate slide generation from any public webpage.

Python Firecrawl API REST APIs WebSockets JWT Auth Web Scraping
Fatigue Sense

Real-time driver safety system unifying a Keras CNN with MediaPipe Pose and Hand tracking across 3 concurrent detection pipelines, flagging fatigue, posture, and hand-position risks with 95% classification accuracy.

Python Keras MediaPipe TensorFlow OpenCV Pygame
AgroDoc

Full-stack plant disease detection web app using a TensorFlow CNN served via Streamlit, classifying 38 conditions across 14 crops at 97.38% accuracy on 87,000+ images.

Python TensorFlow Keras OpenCV Streamlit
Publications

Published work

Research writing connected to applied AI, agriculture, and computer vision.

Advances in Plant Disease Detection and Classification Systems
Scrivener Publishing LLC · Nov 2025

Co-authored a chapter published in Optimizing AI Applications for Sustainable Agriculture, documenting findings from the AgroDoc research project. The chapter covers CNNs, transfer learning, data augmentation, IoT integration, and precision agriculture applications for plant disease detection.

Deep Learning Computer Vision Sustainable Agriculture AgroDoc
Contact

Let's connect

Have a project in mind or want to collaborate? I'd love to hear from you.

I'm currently open to full-time software engineering roles. Feel free to reach out.

Email
karanbirsinghpelia@gmail.com
LinkedIn
karanbir-pelia
GitHub
karanbir-pelia
Location
Stony Brook, New York