Jaxon Poentis

jaxonp808@gmail.com|linkedin.com/in/jaxon-poentis|github.com/jaximus808|jaxonp.com

Education

Washington University in St. Louis

May 2027

Bachelor of Science in Computer Science and Mathematics | GPA: 3.75/4.0

St. Louis, MO
  • Honors: ISEF Finalist (2022); Dean's List; Eagle Scout | YC Startup School 2026 (SF)
  • Coursework: Intro to AI, Cloud Computing with Big Data Applications, Systems Software, Analysis of Algorithms, Concurrency & Memory-Safe Systems (Grad), Theory of Parallel Systems (Grad)

Experience

Northstar Robotics

Sept. 2026 – Present

Founding Engineer

Remote, SF
  • Shipped a multi-relay supervisor for robot teleoperation and data collection; built reusable installers, configuration tools, and hardware diagnostics to support partner deployments.
  • Built a reliable recording pipeline that raised three-camera capture from 7.5–11.6 to 30 fps; all 29 test recordings reached Azure within 9–36 seconds of stopping, with synchronized video, joints, and actions.
  • Resolved production camera-session leaks and network dropouts; added automatic reconnection and recording-upload recovery after restarts to support remote operators.
  • Built a Python/Modal pipeline to prepare robot demonstrations for policy fine-tuning on A100 GPUs, synchronizing video and actions, rejecting incomplete episodes, and holding out sessions for validation.

Capital One

June 2026 – Aug. 2026

Software Engineer Intern

McLean, VA
  • Built a FastAPI app on a 5-person team to replace manual CIDR and rule lookups, using PostgreSQL GiST indexes for subsecond searches across 1.5M+ security groups and rules.
  • Built an in-app chat integration to eliminate separate API token provisioning by wrapping Claude Code in a driver; adopted by the team with expanding use across the organization.

Tesla

Jan. 2026 – May 2026

Software Engineer Intern

Fremont, CA
  • Led development and deployment of a distributed AI agent platform (Python, FastAPI, LangGraph, MongoDB, WebSockets) connecting 12+ tool interfaces for data retrieval and workflow execution across internal microservices; adopted by billing, support, and application teams.
  • Engineered asynchronous, event-driven workflows across 10+ .NET microservices to orchestrate energy lease transfers and propagate account data in production, serving Tesla's entire U.S. energy customer base.
  • Led a cross-team rollout of account-based authentication across the microservice fleet, integrating Cerberus with on-behalf-of identity propagation and account passthrough for secure, scoped agent tool execution.
  • Profiled and tuned service latency with distributed tracing, APM, and load tests against production SLAs; deployed and managed 10+ services on Kubernetes through Jenkins CI/CD pipelines.

Spectrum

May 2025 – Aug. 2025

Software Engineer Intern

St. Louis, MO
  • Built a Python/AST dependency analyzer, saving 2–4 validation days per refactoring cycle.

Projects

GeonosisAI – Robot Deployment Factory

Oct. 2026

Python, TypeScript, SpacetimeDB, Modal, MuJoCo

  • Honorable Mention, SpacetimeDB track, Cal Hacks 13.0. Built an agent pipeline on SpacetimeDB that evaluates a robot policy, clusters failures with Claude, buys recovery data from contributors' phones, fine-tunes π0.5 on Modal, and re-evaluates before shipping.

Tandem – AI Agent Orchestration

May 2026 – Present

Go, TypeScript, MCP, SQLite

  • Built an agent-orchestration platform with MCP integration, a durable task queue with atomic claims and heartbeat expiry, and a live web UI where users review plans, approve work, and track parallel agent sessions.

Delivery Robot Platform

Aug. 2025 – Present

Go, Kafka, gRPC, WebSockets

  • Lead a 20-person team building a campus delivery platform; developed Go microservices for fleet coordination, a JWT authentication service, and robot command interfaces, with Docker and GCP deployment.

Technical Skills

  • Languages: Python, TypeScript, Go, SQL, C#, C++, Java, JavaScript
  • AI & Applications: LangGraph, MCP, FastAPI, React, Next.js, Node.js, PyTorch
  • Data & Integrations: PostgreSQL, SpacetimeDB, MongoDB, SQLite, WebSockets, Kafka, gRPC, .NET
  • Deployment: Linux, Docker, Kubernetes, Azure, GCP, Modal, Jenkins, GitHub Actions