Anchit Thakur — ML Fundamentals Portfolio

Who I Am

Professional Bio

Education

Indiana Wesleyan University

Master of Science — Artificial Intelligence, Data Analytics (In Progress)

Carnegie Mellon University

Master of Software Engineering — Scalable Systems (Dec 2022)

Jaypee Institute of Information Technology

Bachelor of Technology in Engineering (May 2021)

Current Role

Software Development Engineer II

Amazon — Tax Engine Artisans, Seattle

Leading development of the Tax Auditor North Star project, using LLM and ML techniques to automatically audit tax records and identify mis-charge patterns for the world's largest transactional tax service (millions of requests/minute).

Technical DNA

Java Python TypeScript AWS React Node.js LLMs ML Pipelines DynamoDB CI/CD

I'm a software engineer who thrives at the intersection of scalable systems and machine learning. Currently pursuing a Master of Science in AI & Data Analytics at Indiana Wesleyan University, I'm deepening my expertise in ML fundamentals while bringing years of production experience from Carnegie Mellon and Amazon.

My journey has taken me from building autograder platforms that scale across 27+ universities (SAIL 2.0 at CMU) to architecting tax calculation APIs handling millions of requests per minute at Amazon. Today, I'm leading projects that leverage LLMs and ML to automate complex audit workflows — proving that AI isn't just a buzzword but a production-grade engineering discipline.

What I Bring

Personal Value Proposition

I turn complex machine learning concepts into production-grade, scalable systems that deliver measurable business impact — combining deep software engineering foundations with a relentless drive to make AI accessible and practical.

01

Systems at Scale

Experience building and maintaining services that process millions of requests per minute. I understand that ML models are only as good as the infrastructure that serves them.

02

ML in Production

Currently applying LLM and ML techniques to real-world audit problems at Amazon scale. I don't just train models — I ship them into production pipelines that matter.

03

Making AI Accessible

From interactive timelines to developer tools, I believe in making complex AI concepts understandable. Contributed to Amazon's internal MCP server improving developer productivity across teams.

My Work

Portfolio Artifacts

Each artifact demonstrates a specific competency with documented objectives, process, tools, and value delivered.

View Artifact →

Interactive Web App

Artificial Intelligence: A Timeline of Milestones

An interactive, scroll-snapping visualization of AI history — from Turing's 1950 paper through AI Winters to the modern ChatGPT era.

Objective Research and visualize the complete history of AI, demonstrating how historical context shapes current ML development

Process Researched primary academic sources (Turing 1950, Dartmouth proposal, NeurIPS papers), curated 15+ milestones, designed era-specific visual theming

Tools HTML5, CSS3, JavaScript, Intersection Observer API, Google Fonts, Cloudflare Pages

Value Demonstrates ability to synthesize complex technical history into an engaging, accessible interactive experience

View Artifact →

Canva Presentation

Transforming Education (ML Adaptive Learning)

An exploration of how Machine Learning and Generative AI can create personalized, adaptive learning pathways to eliminate the friction of static systems.

Objective Explore how ML and Generative AI create personalized, adaptive learning pathways that eliminate the friction of static systems.

Process Collaborated with my team to synthesize ML concepts, analyze learning friction points, and propose AI-driven dynamic course correction.

Tools Canva, Claude 4.6 Opus, Google Gemini Pro

Value Demonstrates the ability to apply ML concepts to real-world educational challenges, translating technical capabilities into strategic leadership solutions.

Coming Soon

Next Artifact

More artifacts will be added throughout the course as new competencies are demonstrated.