Mathematics.
Software Engineering.
AI.

End-to-end solutions for businesses that need
hard technical problems solved — fast.

Explore Services Get in Touch
Case Study

From Zero to
Pitch-Ready in 3 Months.

12 disease prediction models. A complete web application.
Clinical Atlas with SHAP explanations. Solo delivery.

Read the Story
About

25+ Years.
Google. Amazon. PhD.

46 peer-reviewed papers. $100M+ revenue impact.
One SIAM book. One relentless problem solver.

Meet Dr. Livne
About
Dr. Oren Livne
PhD Applied Mathematics Weizmann Institute 25+ Years Experience Who's Who in America

Dr. Oren Livne

I turn ambitious ideas into shipped products. Whether it's an ML pipeline, a full-stack web application, or a pricing engine — I deliver the work of a team, solo, on schedule.

With a PhD in Applied Mathematics from the Weizmann Institute of Science and 25+ years of experience, I combine deep technical skill with the pragmatism to build things that actually work in production. I'm equally comfortable writing numerical solvers, designing React UIs, architecting cloud infrastructure, or presenting to executives.

At Google, I built large-scale optimization algorithms for DoubleClick/DFP that generated over $100M in YouTube ad revenue. At Amazon Lab126, I developed fast algorithms for deep neural network training and compression powering computer vision products. As VP of AI at CulminationBio, I built multimodal search products on the world's largest health data lake and led the engineering team. Currently at QXO, I serve as Principal AI Scientist, building pricing optimization systems for a $7B building products distributor and managing engineering teams.

I've published a book with SIAM on multigrid methods and authored 46 peer-reviewed papers. My open-source contributions include LAMG (algebraic multigrid solver) and PRIMAL (DNA imputation). I'm also a competitive chess player (USCF ~2100) and a Marquis Who's Who in America honoree.

25+
Years of Experience
46
Published Papers
$100M+
Revenue Impact
1
SIAM Book
Services

What I Build

One person. Full stack. From concept to deployed product.

End-to-End Product Development

  • Complete web applications: React frontends, Python/FastAPI backends, PostgreSQL
  • Cloud infrastructure on GCP/AWS — from architecture to deployment
  • Payment integration (Stripe), shipping/fulfillment (EasyPost), admin panels
  • API design, microservice architecture, CI/CD pipelines
  • Product management: requirements, sprint planning, stakeholder communication

AI & Machine Learning

  • ML pipelines: disease diagnostics, risk prediction, clinical data analysis
  • Deep neural network training acceleration and model compression
  • Computer vision, image segmentation, NLP feature extraction
  • Knowledge graphs, network analysis, recommender systems
  • Model explainability (SHAP) and clinical-grade validation

Data & Business Analytics

  • Pricing optimization and revenue modeling
  • Supply chain software, inventory tracking, fulfillment automation
  • Data pipeline architecture and ETL at scale
  • Interactive dashboards and data discovery platforms

Scientific Computing & Applied Mathematics

  • Fast numerical solvers for physics, molecular dynamics, quantum chemistry
  • Algebraic multigrid methods for graph problems in ML/data science
  • High-performance computing and parallel algorithm optimization
  • Computational fluid dynamics, mesh refinement, signal processing
Case Study

Taldavi Health Clarity

From zero to pitch-ready in 3 months.

A clinical AI startup needed a complete disease diagnostics platform — and a consumer product — built from scratch. They needed it fast.

I built an end-to-end disease diagnostics pipeline that predicts disease risk from routine blood tests (CBC/CMP panels). The system includes 12 validated disease models with clinical-grade performance, a novel anti-leakage methodology that prevents diagnostic formula contamination, and an automated "disease factory" that reduces new disease onboarding from months to hours.

Beyond the ML pipeline, I built the full Pathfinder product: a complete React web application with Stripe payments, EasyPost fulfillment, an admin panel, PostgreSQL backend, FastAPI middleware, and beautiful PDF reports for cancer patients.

I also produced the Clinical Atlas — a comprehensive reference with per-disease PDFs, SHAP explanations, and comparisons to state-of-the-art published literature.

Model Performance

Type 2 Diabetes
0.949
AUC
SLE (Lupus)
0.924
AUC
MASH
0.887
AUC
+ 9 More Diseases
12
Total Models
3 months Zero to pitch-ready
Hours, not months To add a new disease
Full product suite Pipeline + webapp + reports
Solo delivery Work of a 6-person team
Contact

Let's Talk About Your Hardest Problems

Whether it's a tough optimization problem, a machine learning pipeline, or a full product build —
I'd love to hear about it.

pinebirchanalytics@gmail.com