Skip to content

Paul Kathmann · Software Engineer & Entrepreneur

I build compute infrastructure & AI Agents for
production workflows.

Software engineer at Palantir, working on compute infrastructure and internal AI agents. Former founder. Five countries, a few products shipped.

About

My Story

I'm a software engineer and former founder. Today I work at Palantir on compute infrastructure and internal AI agents — building the systems and tooling that let agents generate and deploy production workflows in minutes instead of days.

Before Palantir, I founded Fanita Health, a telehealth platform for stigmatised healthcare in Indonesia. We grew it to 50,000 users and a team of ten, and raised a six-figure pre-seed round. That experience shaped how I think about building products — not just as systems, but as tools that solve real problems at scale.

I've lived and worked across five countries. My background spans business, product, and engineering, and I'm most interested in the boundary between agents and the production systems they touch.

Experience

Where I’ve worked.

Jul 2025
Now

Software Engineer · Palantir Technologies

New York, US

  • Backend engineer on the compute modules powering agentic AI and distributed data applications for enterprise customers — focused on scale, fault tolerance, and low-latency execution.
  • Built and deployed an agentic system for backend application development, enabling AI agents to generate, test, and deploy production-ready services end-to-end.
  • Developed a suite of Model Context Protocol (MCP) tools spanning the full compute-module lifecycle: provisioning, execution, validation, iteration.
May 2024
Dec 2024

Software Engineer (Intern) · Pattern Inc.

Utah, US

  • Built a RAG-based AI service to automate content creation and analysis using FastAPI, Snowflake, OpenAI, Pinecone, and DynamoDB.
  • Cut LLM costs 85% by fine-tuning BERT to classify and route prompts to the cheapest viable model.
  • Reduced OpenAI costs 45% by replacing parts of the image-processing pipeline with AWS Rekognition, Airflow, and AWS Lambda.
  • Saved the content team ~10 hours of weekly work by building an auto-translation service on GitHub Actions, FastAPI, and OpenAI.
Sep 2023
Dec 2023

Software Engineer (Research Assistant) · University of Pennsylvania

Philadelphia, US

  • Designed the system architecture for a voice-enabled AI tutor teaching English to students in India.
  • Led the development of a full-stack LLM platform in Python, MongoDB, HTML/CSS, and TypeScript.
  • Reduced chatbot latency 80% via response streaming and prompt engineering.
Jun 2023
Aug 2023

Software Engineer (Intern) · Blockbrain

Hamburg, Germany

  • Led end-to-end full-stack development of an Enterprise AI Assistant using FastAPI, Pinecone, MongoDB, React, and Firebase.
  • Reached 100+ enterprise users via the reliability of the underlying RAG system; the launch's success prompted the company to pivot to focus on it full-time.
Sep 2020
Dec 2023

Co-Founder, CEO/CTO · Fanita Health

Jakarta, Indonesia

  • Founded the business and built the prototype of a telehealth platform for stigmatised healthcare.
  • Led product, marketing, and engineering for a platform that grew to 50,000 active users.
  • Raised a six-figure pre-seed round and recruited a team of ten full-time employees.

Education

Where I studied.

Aug 2023
May 2025

Master in Computer Science · University of Pennsylvania

Philadelphia, US

  • GPA 4.00 / 4.00.
  • President of CIT59x.
  • Coursework: Machine Learning, Deep Learning, Algorithms & Computation, Distributed Systems.

Toolkit

What I work with.

Languages

TypeScriptPythonJavaC / C++GoRubySQLR

AI / ML

PyTorchScikit-learnPandasNumpyLangChainOpenAIPinecone

DevOps & Cloud

DockerKubernetesTerraformAirflowGitHub ActionsAWSGCPFirebase

Web & Data

ReactFastAPIFlaskRuby on RailsMongoDBSnowflakeDynamoDB

Get in touch

Find me online.

If you're working on something interesting at the boundary of AI and production systems, I’d love to hear about it.