Home
Resume ↗

YASH’S PORTFOLIO / QUICK FIND

What are you looking for?

A FEW GOOD PLACES TO START

SEARCH THE WORK, FOLLOW THE EVIDENCE.ESC TO CLOSE

SOFTWARE & AI ENGINEER · SAN JOSE, CA

Yash Kamerkar

Turning complexity into clarity.

I’m Yash, a software and AI engineer at Uber, building dependable backend services and useful AI applications.

Production backend & AI systemsUber · Dell · Universal Courier

Explore my experienceView resume

01 / PROFESSIONAL EXPERIENCE

Built in the real world.

Full resume ↗
CURRENT ROLE

Uber

Aug 2026 — Present

USA

Software Engineer | AI Engineer

Backend engineering · Applied AI

  • Engineered a RAG pipeline with LangChain, OpenAI API, and Pinecone for an internal driver policy assistant, cutting policy lookup time from 15 minutes to under 50 seconds across 12,000 compliance documents.
  • Architected a gRPC trip-event aggregation microservice in Java and Spring Boot, consolidating 4 upstream location feeds and cutting duplicate event processing incidents from 9 to 2 per week.
PREVIOUSLY

Universal Courier

Jul 2023 — Aug 2024

India

Software Engineer | AI Engineer

Conversational AI · Logistics systems

  • Built an LLM-powered support assistant with LangChain and OpenAI API, automating 420 repetitive delivery query resolutions per month and cutting average first-response time from 6 hours to 12 minutes.
  • Designed a scikit-learn delivery ETA model trained on 2 years of shipment records, reducing average estimation error from 3.4 hours to 1.1 hours across 5 regional delivery zones.
4 more contributions
  • Integrated 3 carrier REST APIs (UPS, FedEx, USPS) using Python and FastAPI to consolidate real-time tracking for 8,000 daily parcels into a unified customer-facing dashboard.
  • Constructed FastAPI endpoints for 3 internal logistics dashboards, supporting 150 concurrent dispatcher users during peak morning scheduling windows.
  • Developed a Python, Pandas, and SQLAlchemy ETL pipeline ingesting and normalizing 180 GB of daily shipment records from 6 warehouse management systems.
  • Automated nightly route-optimization jobs with Jenkins and Docker on Microsoft Azure, reducing manual scheduling overhead from 4 hours to 35 minutes per dispatch cycle across 3 distribution hubs.
PREVIOUSLY

Dell Technologies

Sep 2021 — Jul 2023

India

Software Engineer

Microservices · Data infrastructure

  • Implemented Java 11 and Spring Boot 2.x order-management microservices for Dell's supply chain platform, processing 45,000 daily purchase requests across 6 service modules.
  • Optimized 5 PostgreSQL reporting endpoints with composite indexes and 3 N+1 query rewrites, reducing average database response time from 2.1 seconds to 340 ms.
5 more contributions
  • Migrated 3 legacy Oracle PL/SQL batch jobs to Python ETL pipelines with Pandas and SQLAlchemy, reducing nightly processing time from 6 hours to 2.5 hours.
  • Authored JUnit 5 and Mockito test suites for 12 service modules, eliminating 14 critical regression defects before each production release and adding 1,500 net tested lines.
  • Deployed containerized Spring Boot services with Docker on Kubernetes, enabling zero-downtime rolling updates across 4 staging and production environments.
  • Streamlined 3 product catalog API contracts using Spring MVC and Hibernate ORM, cutting internal portal integration onboarding from 5 days to 2 days.
  • Refactored an inventory reporting monolith into 2 independently deployable Spring Boot services, cutting build time from 14 minutes to 5 minutes and removing 3 cross-team deployment dependencies.

02 / SELECTED PROJECTS

Made with curiosity.

All repositories ↗

A few things I’ve put into the world.

Context Window’s actual newsroom replay showing collaborating AI agents

Autonomous AI newsroom

Context Window

Eleven AI agents research, verify, and publish a daily AI briefing for engineers.

  • Python
  • LangGraph
  • MCP
  • Next.js
Live demo GitHub
Bill + EOBRule checksVerified evidence

Agentic bill auditing

Medical Bill Auditor

A LangGraph workflow checks synthetic medical bills against billing rules and explains findings with verified evidence.

  • Python
  • LangGraph
  • MCP
  • RAG
Live demo GitHub
Historical data → prediction model → dashboard

Flight-delay research / CO-AUTHORED

FlightSense

Co-authored research connecting flight-delay prediction, cloud infrastructure, and conversational AI.

  • Python
  • AWS
  • XGBoost
Read paper GitHub Research notes
Mettle’s actual browser benchmark showing its compression study’s speed, accuracy, and instruction-following results

In-browser AI benchmark

Mettle: Test your machine’s mettle

Run a 360M-parameter LLM on your GPU. In a 500-question RTX 4080 Laptop study, 4-bit was 12% faster; instruction-following fell from 48% to 9%.

  • TypeScript
  • React
  • WebGPU
  • WebLLM
Live demo GitHub
Yash Kamerkar outside by the waterfront
Yash Kamerkar · San Jose, California

03 / ABOUT

An engineer.
Always curious.

I enjoy the space between dependable backend systems and useful AI applications.

My work has taken me through enterprise technology at Dell, logistics at Universal Courier, and now engineering at Uber. I like turning complex problems into something people can actually use.

M.S. in Computer Science

Stevens Institute of Technology

Sep 2024 — May 2026 · Hoboken, NJ
More on LinkedIn ↗

04 / TOOLKIT

Tools I work with.

Backend

  • Java
  • Spring Boot
  • Python
  • FastAPI

APIs, microservices, and data pipelines.

Applied AI

  • LangChain
  • LangGraph
  • PyTorch
  • scikit-learn

Retrieval, language models, and ML.

Infrastructure

  • AWS
  • Azure
  • Docker
  • Kubernetes

Cloud infrastructure and CI/CD.

View the complete toolkit

Languages

  • Python
  • Java
  • JavaScript (ES6+)
  • TypeScript
  • SQL
  • Bash

AI/ML & Generative AI

  • LangChain
  • LangGraph
  • OpenAI API
  • Anthropic Claude API
  • Retrieval-Augmented Generation (RAG)
  • Large Language Models (LLMs)
  • Prompt Engineering
  • Hugging Face Transformers
  • BERT
  • scikit-learn
  • PyTorch
  • Vector databases

Frameworks & Libraries

  • FastAPI
  • Flask
  • Spring Boot
  • Spring MVC
  • Hibernate ORM
  • Pandas
  • NumPy
  • React.js
  • JUnit 5
  • Mockito

Databases

  • PostgreSQL
  • MySQL
  • Oracle
  • MongoDB
  • Pinecone
  • ChromaDB
  • Redis

Cloud & DevOps

  • Amazon Web Services (AWS)
  • Microsoft Azure
  • Azure OpenAI Service
  • Docker
  • Kubernetes (K8s)
  • Jenkins
  • GitHub Actions
  • Terraform
  • CI/CD Pipelines

Tools & Methodologies

  • Git
  • GitHub
  • JIRA
  • Postman
  • Maven
  • Gradle
  • Linux
  • REST APIs
  • Microservices Architecture
  • Agile/Scrum

05 / CONTACT

Have something in mind?
Let’s talk.

GitHub ↗LinkedIn ↗Twitter / X ↗Download resume ↓