Software Engineer

Abhinav Yedla

Engineer. Builder. Deep in distributed systems.

Software Engineer II at Booz Allen Hamilton. Building cloud infrastructure and AI tooling. Starting Georgia Tech OMSCS Fall 2026.

View My Work Resume

About

A bit about me

I'm a software engineer with two years of full-time experience at Booz Allen Hamilton, working across federal cloud infrastructure, AI systems, and test engineering. The work has been varied — from FAFSA AI prototypes and Kubernetes migrations to building accessibility pipelines and pushing back on AI governance proposals that didn't hold up to scrutiny.

Outside of work, I spent the last year building VTaaS from scratch — a production-deployed distributed video transcoding service on AWS. Not a tutorial project, not a toy app. Real cloud infrastructure: presigned uploads, SQS-decoupled workers, ECS Fargate containers, CloudWatch debugging, and IAM policy simulation to root-cause failures that don't surface in local dev. I needed something that would force me to think like an infrastructure engineer, and it delivered.

I also shipped 2llm, a live AI tooling product that converts ChatGPT memory exports into Claude-compatible formats. It picked up 8,500+ organic views on Reddit at launch — useful signal that the problem was real and the solution resonated.

In Fall 2026, I'll be starting Georgia Tech's OMSCS program. I'm aiming for senior backend and infrastructure roles at top-tier companies — the kind of work where distributed systems depth actually matters.

Experience

Where I've worked

Booz Allen Hamilton·Software Engineer II

July 2024–Present

DOE — FAFSA AI Assistant

—Won internal AI innovation competition (200+ attendees, senior leadership panel); awarded dedicated engineering hours.
—Built React + GPT-3.5 prototype; designed RAG architecture using LangChain for FAFSA documentation retrieval. Development paused due to DoE contract restructuring.

DOE — Cloud Infrastructure

—Contributed to Kubernetes-based cloud migration for Department of Education workloads across containerized microservices.

OTCnet — AI Innovation & Governance

—Leading proposal to fiscal services for AI integration across OTCnet's test engineering organization.
—Architected three-tier accessibility pipeline: Lighthouse, axe-core + Playwright, and LLM-assisted semantic analysis — with hard constraint that the LLM makes zero compliance decisions.
—Authored governance framework covering data handling tradeoffs, PII exposure protocols, human review requirements, and shared prompt templates.
—Challenged leadership's 40–60% time savings projection with grounded cost-benefit analysis; reframed initiative around sustainable adoption.
—Evaluating local LLM vs. cloud API tradeoffs for FedRAMP authorization pathway.

OTCnet — Test Engineering

—Own manual and automated test coverage for OTCnet, the CitiBank federal payment platform.
—Expanding automation suite with Python, JavaScript, Selenium, and Playwright; implementing semantic locators and self-healing scripts.
—Performance testing via Azure Load Testing and AI-assisted JMeter script generation.

Booz Allen Hamilton·Software Engineering Intern

May–August 2023

—Built automated federal procurement data pipeline (Python + USASpending API); reduced retrieval from days to minutes.
—Rebuilt TFX internal platform with React/Flask full-stack overhaul; enabled vendor onboarding across federal agencies.
—Conducted procurement data analysis using regression and clustering to surface agency purchasing patterns.

MetLife·Software Engineering Intern

May–August 2022

—Built JavaScript/PowerShell automation scripts for server restart workflows across production applications.
—Managed AWS EC2 documentation platform; automated scheduled S3 backup pipeline.
—Resolved bottlenecks in 1Desk Automation platform, reducing processing time on high-frequency workflows.

USDA APHIS·Research Intern

May–December 2021

—Built Python/R dispersal model to optimize citrus disease eradication strategy; findings adopted across US growing regions.

Projects

Things I've built

VTaaS

Video Transcoding as a Service

NestJSAWS ECS FargateS3SQSPostgreSQLDockerGitHub Actions

Production-deployed async video transcoding service on AWS. Presigned S3 uploads, SQS-decoupled worker, ECS Fargate containers, Neon Postgres job state, and a CI/CD pipeline with SHA-tagged ECR deploys.

Diagnosed in production

—AWS SDK CRC64NVME checksum behavior change silently broke all S3 downloads — masked by admin credentials in local dev, root-caused through CloudWatch logs.
—S3 returning 403 instead of 404 on HeadObject due to missing s3:ListBucket on the ECS task role — diagnosed via IAM policy simulation.

Live Demo GitHub

2llm

AI Memory Migration Tool

Next.jsVercel Edge FunctionsAnthropic APIUpstash Redis

Live tool that transforms ChatGPT memory exports into structured Claude-compatible formats. BYOK authentication with zero server-side key storage — API key lives in sessionStorage only, cleared in the finally block of every request. Picked up 8,500+ organic views on Reddit at launch.

Live App GitHub

Orai

Mental Health Assessment Tool

React

Behavioral screening app assessing Depression, Anxiety, OCD, ADHD, and PTSD using a condition-weighted scoring algorithm. Built in collaboration with mental health professionals; validated diagnostic alignment with affected users.

GitHub

Skills

What I work with

Languages

PythonTypeScriptJavaScriptJavaC/C++BashPowerShellR

Web & Backend

ReactNext.jsNestJSNode.jsExpressFlaskREST APIsPrisma

Cloud & Infrastructure

AWS (ECS Fargate, S3, SQS, ECR, IAM, SSM, CloudWatch)AzureDockerKubernetesTerraformLinux

Databases

PostgreSQLMongoDBRedis

Data & ML

NumPyPandasTensorFlowPyTorchLangChain

Testing & DevOps

SeleniumPlaywrightJestVitestJUnitTestNGGitHub ActionsCI/CDAzure Load Testing

AI

Anthropic APIOpenAI APIVercel AI SDKPrompt EngineeringLLM Governance

Let's talk.

Open to backend, infrastructure, and distributed systems roles. Reach out directly.

abhinavyedla02@gmail.com linkedin.com/in/ayedla github.com/abhinavyedla02