Skip to content
Nitheesh Kodarapu

Open to software engineering and ML roles

Hi, I'm Nitheesh.
I build software people actually use.

Software engineer and CS student at Texas A&M. I've built AI evaluation at IBM, modernized a 20-year-old point-of-sale system at Waste Management, and made a course planner that 900+ Aggies use.

Resume Email LinkedIn GitHub
Portrait of Nitheesh Kodarapu
hi there!

currently…

  • researching

    Energy-efficient LLM serving, as an undergraduate researcher at Texas A&M

  • maintaining

    AggieSB+, the course planner 900+ Aggies use to pick classes

  • studying

    B.S. in Computer Science (May 2027)

01 / experience

What I've built at workclick the pictures!

May 2026 – Aug 2026

IBM

Software Engineer Intern · watsonx Orchestrate

Built the evaluation pipeline that picked the retrieval backend for IBM's agentic AI platform, plus internal tools for triage and release quality.

hallucinations after the team adopted the top backend
−35%
hallucinations after the team adopted the top backend
citation precision
94%
citation precision
GitHub issues synced into a filterable view
65k+
GitHub issues synced into a filterable view
  • Python
  • NLI models
  • Embeddings
  • Django
  • Next.js
  • IBM DB2
Read the case study
Each sentence of a generated answer is checked against its retrieved sources and marked supported, unsupported, or contradicted. Illustrative recreation, synthetic data.

May 2026 – Present

Texas A&M University

Undergraduate Researcher · Efficient LLM Systems

Co-authored a paper on LLM quantization and energy efficiency, now under review at ACM SIGMETRICS. I wrote the benchmarking software and ran every experiment.

paper under review
SIGMETRICS
paper under review
  • Python
  • PyTorch
  • LLM quantization
  • Energy measurement
MODEL × PRECISIONFP16INT8INT4Same workloadautomated runsPower meterenergy per tokenResults will appear here once the paper is published.
The benchmarking harness: every model and quantization level runs the same workload while power is measured. Illustrative recreation, synthetic data.

May 2025 – May 2026

Waste Management

  • Software Engineer (Part-time) · Aug 2025 – May 2026 · Promoted from intern
  • Software Engineer Intern · May 2025 – Aug 2025

Started as an intern modernizing a 20-year-old point-of-sale system used at 200 scale houses, then was promoted to part-time to fix a long-standing memory leak and automate manifest processing with Amazon Bedrock.

RAM after fixing a memory leak
−40%
RAM after fixing a memory leak
operator screens redesigned
30+
operator screens redesigned
manifest processing turnaround
1 wk → 2–3 days
manifest processing turnaround
  • C#
  • .NET
  • Visual Studio Profiler
  • Amazon Bedrock
  • SQL
Read the case study
Memory over a shift. Before the fix, usage kept climbing the longer the app ran; after, it levels off. Illustrative recreation, synthetic data.

Jun 2024 – Jul 2024

Cotton Holdings

Software Engineer Intern

Built a map that overlays live NOAA weather events on customer locations, so sales could call affected customers before those customers called them.

impacted customers reached within 24 hours of a weather event
1,200+
impacted customers reached within 24 hours of a weather event
  • React
  • Python
  • NOAA weather data
  • Mapping
Read the case study
A storm warning area over customer locations. Everyone inside it lands on the sales team's call list. Illustrative recreation, synthetic data.

02 / projects

Things I've built for fun

Aggie Schedule Builder Plus screenshot

Aggie Schedule Builder Plus

A course-planning platform and Chrome extension that puts grade history, professor reviews, and open-seat alerts inside Texas A&M's registration page.

students using it
900+
students using it
grades from 14 years
5.7M+
grades from 14 years
reviews summarized
90k+
reviews summarized
  • Python
  • FastAPI
  • PostgreSQL
  • LLMs
  • Next.js
  • Chrome Extension
Learn moreLive site

Parallel Graph Crawler

A multithreaded C++ BFS crawler using lock-free atomic bit-arrays and batched queues, kept correct across 5,000 concurrent threads.

nodes / sec
28M
nodes / sec
nodes processed
1B+
nodes processed
  • C++
  • Win32 API
  • Multithreading

Raspberry Pi Smart Home Hub

My apartment's fan, LED strips, and lamp all spoke different radios. A Raspberry Pi now talks to the fan over sub-GHz RF with a CC1101, the strips over Bluetooth, and the lamp over Wi-Fi, so one tap controls them all together.

radios behind one control
3
radios behind one control
devices kept in sync
4
devices kept in sync
  • Python
  • Raspberry Pi
  • CC1101 RF
  • Bluetooth LE
  • Wi-Fi (Tuya)
  • Homebridge

Blackjack AI

A custom CNN that recognizes playing cards in real time and recommends the optimal blackjack move.

accuracy
96%
accuracy
on a commodity CPU
60 FPS
on a commodity CPU
  • Python
  • PyTorch
  • OpenCV
Live demoGitHub

More things I've made some from high school!

HealthGPT

An AI assistant that helps doctors reason through a patient's symptoms, built to cut burnout and missed diagnoses.

Next.js · OpenAI API

LiveGitHub

Galaxy Adventures

A space tourism booking site that placed 4th nationally at the 2023 TSA competition.

Next.js · Tailwind

Live

BetterHomeaccess

A faster grade checker for Texas school districts, with a grade-goal calculator and GPA forecaster.

Python · Flask · BeautifulSoup

LiveGitHub
HomeAccessCenter API screenshot

HomeAccessCenter API

A Go API that serves student grades, assignments, and attendance as clean JSON.

Go · REST

DocsGitHub

Settlers of Catan

The board game in the browser: hex board rendering, dice, resources, and turns.

JavaScript · Canvas

GitHub

Forbidden Island

The cooperative board game, rebuilt as a JavaFX desktop app.

Java · JavaFX

GitHub

03 / about

A bit about me

I like building software that makes my life, and my friends' lives, a little easier. Sometimes it's a full platform like AggieSB+, sometimes a quick automation script, sometimes a weekend spent poking at a new AI model just to see how it works.

I'm curious by nature, so I like taking things apart. Once that meant jailbreaking an Amazon Echo Show to run LineageOS and my own bedside dashboard. Away from a screen, I'm usually outdoors or trying something I haven't done before.

education

Texas A&M University

  • B.S. Computer Science (Honors), Minor in Statistics

    Expected May 2027

GPA 3.85 / 4.0

Undergraduate Research Scholar (senior thesis on efficient LLM systems)

Coursework: Design & Analysis of Algorithms, High Performance Computing, Operating Systems, Distributed Systems & Networking, Machine Learning, Reinforcement Learning, Probability Theory

04 / contact

Let's talk

The fastest way to reach me is email. I'm always happy to walk through any of this work in more detail.

nckodarapu@gmail.com LinkedIn GitHub