Engineer & Founder — Seattle, WA

Abaho
Katabarwa

Systems engineer and founder building AI-powered tools that make complex technology accessible to everyone. Previously at Microsoft, Stanford, and Brown.

Download Resume Get in Touch

Current

Founder & CEO — LLMGraph

Previously

Senior SWE at Microsoft (Copilot)

Specialties

Fullstack · Systems Engineering · LLM/AI Platforms

Education

MS Brown University · BS Stanford University

Languages

Rust · Scala · C++ · TypeScript · Python · Java

01

About

Abaho Katabarwa

I'm a Ugandan-American engineer and founder with a deep interest in building products that make powerful technology accessible — not just to developers, but to everyone.

My career spans embedded systems at Bowers & Wilkins (shipping real-time audio infrastructure under sub-10-microsecond latency budgets), conversational AI tooling at Microsoft Semantic Machines, and the evaluation pipeline powering Microsoft Copilot.

Today I'm building LLMGraph — a no-code platform for customer support chatbots and AI agents that turns a team's documentation into grounded assistants and deploys them as REST APIs or embeddable chat widgets, without writing a single line of code.

Outside of engineering, I'm a long-distance runner chasing my next mile personal record.

Based in Seattle, WA
Focus AI Platforms, Systems Engineering
Stack Rust, Next.js, TypeScript, C++, Scala, Java, Python
GitHub abahocodes
LinkedIn abaho-katabarwa
Email
Phone
02

Experience

Sep 2024 — Present

Katabarwa Labs — LLMGraph

Seattle, WA

Founder & Engineer

  • Architected and built llmgraph.ai — a multi-tenant no-code platform for creating customer support chatbots and AI agents grounded in an organization's documentation, deployed as REST APIs and embeddable chat widgets
  • Delivered sub-5-second end-to-end workflow completion with high availability, routing each workflow across all major LLM providers based on its requirements
  • Engineered a high-performance Rust backend optimized for concurrency, memory safety, and low-latency task dispatch
  • Built a Next.js frontend for workflow authoring, execution, and monitoring; implemented full CI test infrastructure with automated deployment pipelines

Sep 2023 — Sep 2024

Microsoft — Semantic Machines

Seattle, WA

Senior Software Engineer

  • Architected and maintained the evaluation and analytics pipeline for Microsoft Copilot, owning infrastructure design, metrics, and reporting
  • Reduced pipeline runtime from multiple days to a few hours through infrastructure optimization and parallelization
  • Built data tooling and reporting that let data science teams track and validate Copilot quality at scale
  • Designed metric calibration infrastructure to verify that Copilot evaluation metrics measured quality faithfully

Feb 2020 — Sep 2023

Microsoft — Semantic Machines

Seattle, WA

Software Engineer II

  • Designed and delivered the full developer toolchain for a conversational AI SDK: VSCode extension with LSP support, CLI, and documentation
  • Drove enterprise customer adoption by resolving integration blockers and shaping the API surface from real-world usage feedback
  • Built CI pipelines and alerting infrastructure; optimized model training pipelines

Jun 2019 — Aug 2019

Microsoft

Cambridge, MA

Software Engineer Masters Intern

  • Built a mobile app using Microsoft Cognitive Services and neural image transformation that let students create AI-generated art
  • Architected a ReactJS-powered community display exhibited at a major cultural institution

Jun 2015 — Jun 2018

Bowers & Wilkins

Redwood City, CA

Software Engineer

  • Architected and delivered the audio source/sink manager powering the Formation Series — commercially shipped networked wireless speakers — under a strict sub-10-microsecond inter-device synchronization latency budget
  • Developed third-party audio plugins enabling Spotify, Roon, and AirPlay support across the Formation Series ecosystem
  • Built the application layer and SDK for a Broadcom VideoHub set-top box; delivered an IP camera discovery and media streaming engine; developed a hierarchical multi-process persistent data store

Dec 2016 — Jun 2018

StreetCode Academy

East Palo Alto, CA

Lead Teacher & Curriculum Developer

  • Developed and taught a 10-week computer science course for students aged 5–81 in the East Palo Alto community
  • StreetCode provides free technology education to underserved communities — learn more
03

Education

2018 — 2019

Brown University

Master of Science in Computer Science

2011 — 2015

Stanford University

Bachelor of Science in Computer Science

04

Projects

Deep Learning · 3D Vision

ShapeNet 3D Style Transfer

Using CycleGAN to learn bijections between 3D object taxonomies in the ShapeNet database — exploring what a car looks like as a plane, or furniture translated across categories.

Read Report →

Robotics · Reinforcement Learning

Baxter Solves the Rubik's Cube

Teaching a Baxter robot to manipulate a Rubik's Cube through simulated vision and real-world single-arm manipulation. An exploration of deep learning applied to physical robotics.

Read Report →
05

Skills

Systems & Languages

Rust C++ Java Python TypeScript Scala

Frontend

React Next.js React Native Flutter iOS

Cloud & Infra

AWS Azure CI/CD Distributed Systems

AI & ML

LLM Orchestration PyTorch TensorFlow Agentic Systems

Embedded & Real-Time

SDK Development Set-top Box Audio Middleware Low-latency Systems

Tooling

LSP VSCode Extensions CLI Tools Testing Infra
06

Resume

PDF preview not available in this browser

Download Resume

resume.pdf · scroll to read

Download ↓

07 — Contact

Let's Talk

Whether you're interested in LLMGraph, have a technical problem worth solving, or just want to connect — my inbox is open.

GitHub LinkedIn