GTM and Sales Engineer @ Navigara | UC Berkeley Statistics
Theo Douwes
Statistics graduate building GTM systems, underwriting tools, and probabilistic decision software.
Outreach automation for GTM teams, quantitative multifamily underwriting, and risk-aware decision frameworks — published metrics only.
About
Probabilistic systems for GTM, underwriting, and risk
Theo Douwes is a UC Berkeley Statistics graduate (B.A., 2019–2023) based in San Francisco. He builds outreach automation for GTM teams, quantitative underwriting tools for multifamily real estate, and probabilistic decision systems spanning Bayesian/MLE inference, prediction-market pricing, and risk-aware decision frameworks.
He taught 400+ students as Head Instructor of Berkeley’s largest DeCal (STAT 198), presented research on algorithmic bias in hiring at Oxford’s Map the System competition, and managed operations for 30 rental units while modeling multifamily acquisitions totaling $5.88M.
Interests include meditation, travel, and practical AI/ML applications. Skills marked as basics or concepts are adjacent and marketable — not claims of years of production ownership.
$5.88M
Multifamily acquisitions structured
$350K+
Purchase-price savings via CMA + models
400+
Students taught as STAT 198 Head Instructor
200K+
TEDx YouTube views as lead organizer
Skills
Keyword surface & probabilistic toolkit
Role keywords plus documented languages, stats methods, underwriting, and GTM tooling. Items marked basics/concepts are adjacent skills — not multi-year production ownership claims.
Roles & focus
- Quantitative Analyst
- Software Engineer
- Data Analyst
- Data Scientist
- GTM / Sales Engineering
- Full-Stack Analytics
- Quant Research
- Risk
- BI
- Statistical ML
Languages & tooling
- Python
- R
- SQL
- Bash / shell
- JavaScript (basics)
- TypeScript (basics)
- Markdown
- Git / GitHub
- Linux CLI
- VS Code / Cursor
- Excel
Python data / ML
- pandas
- NumPy
- SciPy
- matplotlib
- seaborn
- Plotly (basics)
- scikit-learn
- statsmodels
- Jupyter
- requests
- pydantic (basics)
- pytest (basics)
LLM / AI engineering
- OpenAI API
- Anthropic API (basics)
- LangChain (basics)
- LlamaIndex (concepts)
- Prompt engineering
- Embeddings & RAG (concepts)
- Vector search (concepts)
- Hugging Face (basics)
- Structured outputs
- Tool / function calling (concepts)
- Streamlit AI demos
Quant / stats methods
- Probability
- EV / variance / utility
- MLE
- Bayesian inference
- GLM / regression
- Logistic regression
- Monte Carlo
- Bootstrap / resampling
- Hypothesis testing
- A/B testing (concepts)
- Time-series (basics)
- Confidence intervals
- Experimental design
- Feature engineering (basics)
- Kelly / bankroll thinking
Markets / risk / underwriting
- Fat tails
- Blended / realized vol
- Momentum signals
- Prediction-market microstructure
- Backtesting (concepts)
- Position sizing
- Drawdown controls
- Game theory
- Sharpe-style risk/reward
- CMA
- DCF-style cash flows
- Amortization
- Cost segregation
- Interest-rate scenarios
- Hold / sell analysis
- P&L
Data apps / full-stack lite
- Streamlit
- R Shiny
- Excel modeling / VBA / Pivot
- REST APIs (JSON / CSV / HTTP)
- HTML / CSS (basics)
- FastAPI (basics)
- KPI dashboards
- ETL / data cleaning / validation
Cloud, automation & ops
- AWS
- DigitalOcean
- SSH
- cron
- Docker (basics)
- Outreach automation
- Reproducible experiment repos
- CI-friendly Git hygiene
Domains
- SaaS pipeline tooling
- Multifamily underwriting
- Portfolio operations
- Prediction markets
- Algorithmic bias research
- Probabilistic curriculum design
Work · Selected chapters
Experience
GTM engineering, multifamily underwriting, and independent quant practice — each chapter built on documented outcomes.
GTM and Sales Engineer
Navigara
- Designed and implemented shared outreach automation for SDR, marketing, and sales — sequencing, handoffs, and stage visibility.
- Documented playbooks, stage definitions, and messaging templates so workflows were repeatable.
- Supported Mag 7 lead development using industry connections, research, and outreach tooling (no claimed closed deals).
- Advised GTM strategy from competitor-landscape trade-offs; triaged tooling issues cross-functionally.
Quantitative Real Estate Analyst
Piedmont Realty LLC
- Structured and negotiated 2 multifamily acquisitions totaling $5.88M; $350K+ purchase-price savings via CMA and Excel financial modeling.
- Built Excel + R Shiny holding-period models (rates, amortization, cost segregation, taxes, appreciation) plus owner-ready underwriting packs.
- Owned P&L for 30 rental units — tenants, owners, contractors, municipal compliance — and maintained high occupancy.
- Tracked rent rolls, tickets, and contractor spend in Excel; translated models into plain-language owner updates.
Quantitative Analyst
Independent Practice
- Built proprietary analytics software using MLE and regression for behavioral patterns from small samples.
- Applied Bayesian updating for sparse data with documented assumptions; used Git/GitHub for reproducible experiments.
- Applied game theory and risk-management rules (position sizing, stop criteria) in probabilistic live settings; funded university expenses systematically.
Education · Foundation
University of California, Berkeley
Bachelor of Arts — Statistics · Aug 2019 – Dec 2023
- STAT 198 Head Instructor — Poker Theory and Fundamentals; 400+ students across 3 semesters (Spring 2022 – Fall 2023); largest DeCal at UC Berkeley
- Map the System, Oxford University (Summer 2022) — “Algorithmic Bias in Online Hiring Systems”
- Berkeley Venture Capital Group — startup evaluation and outreach
Projects
Selected systems
Bayesian Inference System
Sparse-data decisionsProfiling tool for sparse-data decisions using Bayesian methods — assumptions documented, experiments reproducible.
Real Estate Analysis Tool
R ShinyInteractive R Shiny property investment model with tunable financial parameters for holding-period underwriting.
ZeroCopy.systems Prediction Market Pricing
15-minute binary marketsPricing for 15-minute binary outcome markets with fat-tailed adjustments, blended volatility, momentum signals, and systematic position sizing.
TEDx Lead Organizer
May 2018Led application, speaker selection, and production for an official TEDx event — 200K+ YouTube views.
Writing
Notes & frameworks
Connect