Jacob
Warthen
Building AI systems that protect people, optimize industries, and create new markets. From distributed trading infrastructure to scam prevention at scale.
Versatile by design.
Technical by nature.
At 24, I have spent the last few years building AI systems across petroleum engineering, financial markets, healthcare, and consumer security — not because it was easy, but because I chose the hardest problems available to me at each stage.
I started at Analog Computation Enterprise as an intern while finishing my CS degree in three years at USF. Within months I was the primary developer on production systems, regularly self-teaching entire domains — Modelica, Wolfram Language, digital signal processing, MATLAB — to deliver results no one else on the team could.
After ACE I went deep on algorithmic trading, building a full live-trading infrastructure in C++ and Python — co-located servers, custom backtesting, real-time signal processing — before being pulled into Jump Partners as co-founder and chief engineer to build the first vertical, MyProtection.AI.
I work 80–100 hour weeks because the problems are that interesting. The goal is to build something that genuinely changes how people interact with technology — starting with AI that protects them.
Where the work
has taken me.
Leading all software development across multiple AI-driven verticals under the Jump Partners holding company. Primary focus on MyProtection.AI — a scam prevention platform with 200+ users — and XEO Labs, an AI search optimization product. Responsible for full-stack architecture, cloud infrastructure, and technical strategy across both ventures.
Joined as an intern during the final year of my CS degree and grew to lead the most complex client-facing projects. Built production AI systems across petroleum engineering, healthcare, and financial markets using the proprietary Octopus Neural System (ONS) — a sensor-based AI grounded in Nash equilibrium theory rather than traditional deep learning.
Founded and fully automated an email marketing agency while completing my CS degree at USF. Built custom scripting and software pipelines that ran the business end-to-end with minimal manual input — client meetings were the only thing I could not automate. Profitable, but too small-ceiling to keep my attention.
Systems built.
Problems solved.
Each project required learning a new domain from scratch and delivering production-grade results.
End-to-end scam prevention platform with Gmail, Outlook, and IMAP integration. 15-node AWS EKS cluster, Kafka event streaming, federated learning loop enabling on-device model updates without raw data leaving users' phones. Real-time risk scoring on every inbound email.
Live quantitative trading system detecting non-statistical anomalies — mini "black swan" events — in U.S. equities. Full alpha → risk → execution pipeline, custom backtesting engine, and Prometheus/Grafana monitoring. Live market data streamed via Databento, trades executed through IBKR.
Built a full gas-lifted oil well system simulator from scratch — including a custom oil-gas-water medium and VLP/IPR curve interpolation. Solved the global non-linear optimization problem using ONS. Presented findings at the Wolfram Conference 2023 to an audience of scientists and engineers.
Optimized, co-located HFT system written entirely in C++ targeting sub-microsecond execution. Custom backtesting engine with realistic market simulation. Confirmed exploitable statistical edges; infrastructure costs ultimately made the economics unfavorable at the time.
Designed a personalized dosing system for PRP and bone marrow injections — a largely unregulated space. Used ONS to cluster patients by signal similarity, extracted non-statistical anomalies via DSP, and built an optimization algorithm to recommend optimal viable doses based on baseline patient characteristics.
Co-leading the second Jump Partners vertical — an AI-first search optimization product targeting businesses that need visibility in generative AI responses. Full technical build-out: site analysis engine, reporting pipeline, and client dashboard. Phase 1 live at xeolabs.ai.
Technical depth
across the stack.
- Python
- C++
- Rust
- TypeScript / JavaScript
- Mathematica / Wolfram
- MATLAB
- Modelica
- Go
- Kubernetes / EKS
- Docker
- Apache Kafka
- Apache Flink
- PostgreSQL
- AWS (EKS, S3, CloudFront)
- Firebase
- Prometheus / Grafana
- Federated Learning
- Digital Signal Processing
- Optimization Theory
- Financial Mathematics
- Non-linear Optimization
- Nash Equilibrium AI
- HFT Architecture
- React Native
- React / Vite
- System Architecture
- Cloud-native Design
- Real-time Streaming
- Backtesting Engines
- API Design
Recognition &
milestones.
Delivered a 30-minute technical presentation on gas-lifted petroleum well simulation — covering the physics model, automated generation of custom well systems, and ONS-based global optimization. Presented to scientists and engineers from across the Wolfram research community.
Completed a four-year computer science degree in three years while working a technical internship, founding JDUB Digital, and playing Division I collegiate soccer. Graduated May 2023.
Competed as a Division I collegiate soccer player at USF while simultaneously managing a full technical workload, running a business, and beginning my career in AI. A formative lesson in managing intensity across multiple fronts.
At 24, co-founder and CTO of a holding company with two live AI products, 200+ users, institutional investor attention, and a SOC 2 Type 2 certification in progress. The pipeline of what comes next is the most interesting part.
Let's build
something hard.
If you are working on a genuinely difficult technical problem — in AI, finance, infrastructure, or anything in between — I would like to hear about it.