Abhishek Raju

Hello! 👋

I'm Abhishek Raju

AI Engineer

About Me

AI Engineer specializing in RAG systems, LLM integration, and production generative AI. Architected pipelines processing 50k+ customer reviews (Pinecone + LangChain + OpenAI), built compliance automation scanning 4,800 pages (Playwright + Bedrock), and deployed vision-based AI assistants serving 40+ users. Currently building Chitram (AI image hosting) and Gremlin (AI QA agent with 70+ risk patterns). Strong in Python, AWS AI/ML stack, prompt engineering, and distributed systems. Seeking AI Engineer roles.

Python LangChain RAG Pinecone ChromaDB AWS Bedrock OpenAI Claude API Prompt Engineering Hugging Face FastAPI PostgreSQL Redis Docker Kubernetes AWS (Lambda, Step Functions, SageMaker, ECS) CI/CD CloudWatch

Projects

01

Chitram (చిత్రం) - Image Hosting Service

Image hosting service MVP for learning distributed systems. Built with FastAPI, PostgreSQL, MinIO, and Supabase auth. Features a pluggable authentication system, async thumbnail generation, and comprehensive test architecture.

FastAPIPythonPostgreSQLSupabaseDocker
02

Gremlin — AI Risk Critic (CLI · Python Library · Dashboard)

Pre-ship risk critic that identifies breaking scenarios before code reaches production. 107 domain-specific QA patterns across security, concurrency, payments, and auth. 90.7% tie rate with baseline Claude Sonnet in A/B evals across 54 test cases. Published to PyPI with CLI, programmatic Python API, async support, and interactive Risk Dashboard with heatmap visualizations.

PythonClaude APIPrompt EngineeringCLIGitHub ActionsChart.js
03

Argus - AI Accessibility Assistant

An intelligent accessibility testing tool that leverages AWS Bedrock and custom prompt engineering to identify WCAG compliance issues across web and mobile platforms. The system analyzes page structure semantically, reduces false positives by 40%, and provides actionable remediation guidance for development teams.

AWS BedrockPythonLangChainPlaywrightAccessibility
04

Compliance Automation Pipeline

Distributed web crawling infrastructure (Playwright + AWS ECS Fargate) processing 4,800 pages across 8 locales with event-driven messaging (SNS). Integrated AI-powered anomaly detection cutting manual review from 3 months to 0.5 days.

PlaywrightAWS ECSSNS/SQSGrafanaCloudWatch

Experience

Senior Software Engineer

Amazon

Mar 2021 - Present
  • Architected production RAG pipeline (Python + FastAPI + Pinecone) processing 50k+ customer reviews with LangChain orchestration. Integrated OpenAI and AWS Bedrock with CI/CD deployment, monitoring via CloudWatch for latency/accuracy tracking. Enabled natural language querying surfacing 100+ UX issues, driving 20% support ticket reduction.
  • Co-led Argus, an AI accessibility assistant utilizing custom prompt engineering on AWS Bedrock; integrated multi-platform (Web/Mobile) testing strategies and streamlined defect identification and resolution
  • Designed and deployed generative AI assistant using AWS Bedrock (Claude Sonnet) and Kendra for semantic search across 10k+ WCAG compliance guidelines. Integrated LLM via REST API into partner applications, enabling natural language querying and automated validation workflows. Served 8 engineering teams (40+ users), reducing onboarding from 2 weeks to 1 hour.
  • Directed an 8-month European Accessibility Act (EAA) compliance program for Ring Ecommerce, coordinating 8 teams to build an automated web scanning system across 4,800 pages in 8 locales
  • Engineered an automated testing system for e-commerce workflows that increased test coverage from 5% to 70%, supporting reliable Black Friday launches
  • Managed complete QA lifecycle for Ratings & Reviews launch on 400 Ring.com product pages, achieving zero production defects
  • Oversaw QA for Alexa Personality in 'Hey, Disney!' launch, building automation framework for hundreds of interactive voice responses on 100M+ Echo devices

Software Engineer (Automation)

Veracode

Feb 2015 - Feb 2021
  • Led comprehensive QA for application security product migration from data center to AWS/Kubernetes for 1,000+ enterprise customers
  • Automated cloud infrastructure workflows using BASH, reducing manual intervention by 80%
  • Developed API automation framework achieving 100% test coverage across 4 backend services, reducing regression testing time from 2 days to 2 hours
  • Enhanced UI test coverage from 10% to 75% for application security product through CI/CD integration

Software Intern

EMC (acquired by Dell)

Jan 2013 - Jul 2013
  • Debugged fault containment path in VNX/VNXe file system recovery components and wrote unit tests for FSCK recovery tool
  • Performed system-level programming in C, C++ on BSD UNIX (RTOS) platform and optimized code to reduce API call hits to physical disk

Education

Master of Science, Computer Science

Northeastern University, Boston

May 2014

    Blog

    Thoughts on testing, automation, and software quality.

    February 23, 2026

    Why I Built a Pre-Ship Risk Critic

    The origin story of Gremlin - an AI tool that systematically asks 'what if?' questions before code ships. How 93 curated risk patterns from real incidents add genuine signal on top of LLM reasoning.

    AIQAGremlinPrompt Engineering
    January 19, 2026

    Deployment Debugging: How Code Duplication Caused a Production Bug (Part 3 of 3)

    Deployed automatic AI tagging expecting smooth rollout. Reality: 5 cascading infrastructure bugs, 3 hours of debugging. The final bug? Code duplication caused storage backend mismatch.

    deploymentdebuggingdry-principleproduction-bugs
    January 18, 2026

    AI Tagging Evolution: Manual First, Automatic Second (Part 2 of 3)

    We started with a manual /ai-tag endpoint before building automatic tagging with Celery. Why? Validate the hard part (AI integration) before adding distributed systems complexity.

    aimanual
    January 17, 2026

    Building an AI Provider System: The Strategy Pattern for Vision APIs (Part 1 of 3)

    Built a swappable AI provider system for automatic image tagging using the Strategy pattern. Switch between OpenAI Vision, Google Vision, or mock providers with a single environment variable - zero code changes.

    aistrategy-patternarchitectureopenai
    January 12, 2026

    AI for Social Impact: Ideas Worth Building

    From potholes to plant disease — real-world AI ideas for social impact that go beyond the hype.

    AISocial ImpactEdge AICivic Tech
    View all posts