MISSIPSSA OUNNAR

EN / FR (fluent) · AR (conversational) · Berber

Summary

Undergraduate pursuing a dual BS in Mechanical Engineering and Computer Science at UC Denver. Self-directed machine-learning and quantitative research since December 2025. Hardware hands-on (CAD, mechatronics, custom dual-5090 workstation built from scratch) combined with software depth — LLM compression research under active patent filing, an autonomous agentic AI system, and a live-funded C# quantitative trading system. Full portfolio and case briefs at mounnar.vercel.app.

Research & Projects

UltraCompress — LLM Compression Research mounnar.vercel.app/projects/ultracompress

Patent pending · method held privately · Python, PyTorch, CUDA, dual-GPU workstation

Apr 2026 – Present
  • Novel neural-network compression method that preserves downstream performance under substantial parameter reduction.
  • Cross-family validation across multiple open-weight language models without per-model tuning.
  • Rigorous evaluation methodology: held-out benchmarks, matched-parameter baselines, bootstrap confidence intervals, pinned reproduction guide.
  • Method, specific metrics, and source held privately during active patent filing. Available for NDA-scope discussion.

Athena — Autonomous Agentic AI (Sipsa Labs) mounnar.vercel.app/projects/athena

Private · Python, C#, LangGraph, MCP, Docker

Mar 2026 – Present
  • Multi-layer cognitive architecture targeting genuine reasoning and long-horizon memory rather than single-shot LLM generation.
  • Integrated with live data streams through multiple concurrent services for real-time context.
  • Founded Sipsa Labs as parent company for AI and quantitative ventures.

Multi-Instrument Quantitative Trading System mounnar.vercel.app/projects/quant

C#, Python, NinjaTrader 8, Rithmic L2, XGBoost / LightGBM / CatBoost

Dec 2025 – Present
  • Automated multi-instrument futures trading system running NQ, ES, Gold, and ETH concurrently on a live Level-2 CME data feed.
  • Walk-forward machine-learning pipeline with regime-conditional signals, calibrated position sizing, and conservative risk controls.
  • Passed the TopStep $150K Express funded-account evaluation on live data. Specific signals, features, and risk logic held privately.

Custom HPC Workstation — Dual-GPU Liquid-Cooled Build

AMD Ryzen 9 9950X3D · 2× RTX 5090 (liquid cooled) · 128–256 GB DDR5

2025 – Present
  • Designed and assembled a dual-GPU workstation from scratch to run concurrent live trading, ML training, and agentic AI inference workloads. AVX-512-optimized, thermal-tuned.

Motorcycle Engineering Project

SolidWorks · MATLAB · FEA · Mechatronics

2025
  • Custom build applying materials science, torque/stress analysis, and mechatronics. Integrated mechanical and electronic subsystems.

Education

University of Colorado Denver

Expected May 2027

B.S. Mechanical Engineering + B.S. Computer Science
Minor: Physics · GPA: 3.6
Previously pursued Aerospace Engineering.

Self-directed ML & Quantitative Research

Dec 2025 – Present

Market microstructure, neural-network compression, agentic architectures. Produced compression research under active patent filing in five months.

Certifications

  • OSHA 10 · General Industry2020 – Present
  • U.S. Army MOS 91B · Wheeled Vehicle Mechanic2020 – 2021

Experience

Operations Associate

Denver International Airport

2023 – Present

Airfield operations at a major international hub: FAA Part 139 inspections, wildlife monitoring, NOTAM issuance, snow-removal coordination across ATC, airline, and ground teams.

Undergraduate Teaching Assistant

UC Denver — Engineering Fundamentals & Engineering Design

2023 – Present

Support 50+ undergraduate students per semester through labs, office hours, and design-project mentorship.

CAD Drafter

XDS Fire · Denver, CO

2022 – 2023

NFPA-compliant fire-suppression drawings for Fortune 500 and government clients using AutoCAD.

Engineering Intern

Geocal Inc. · Denver, CO

2022 – 2023

Environmental restoration fieldwork on the Little Blue Creek project under a licensed PE, in coordination with the U.S. Fish & Wildlife Service.

Skills & Technologies

Languages
Python · C# · C · C++ · Java · SQL · TypeScript · JavaScript · PowerShell · Pine Script · MATLAB
Engineering
AutoCAD · SolidWorks · ANSYS · FEA · CFD · Mechatronics · Hardware Design
ML / Quant
PyTorch · scikit-learn · XGBoost · LightGBM · CatBoost · pandas · NumPy · statsmodels · walk-forward backtesting · Monte Carlo
Trading
NinjaTrader 8 · Rithmic L2 · TradingView webhooks · Level-2 order flow
Web / Backend
Next.js · React · React Native · FastAPI · Flask · WebSocket · Docker · Tailwind
Data / Systems
PostgreSQL · SQLite · Redis · Chroma · Linux / Windows admin · Git · CUDA · AVX-512 · HPC tuning · MCP servers