Jonathan Gu

I build ML systems that turn predictions into product decisions.

At Instacart, I build ML systems that make decisions: bids, rankings, forecasts, and fulfillment actions. My ads bidding system replaced manual product-keyword bids with weekly-budget optimization. It now covers roughly 70% of Sponsored Products spend, about $700M annualized.

The pattern is simple: start with noisy signals, choose the action, measure the result, and keep the loop safe when real users and money are involved. The ads work is reflected in issued patent US 12,511,677 B2.

Outside work, I build Bountiful Garden, OpenClawBrain, and Project Pelican: products about trust, memory, and decisions under uncertainty.

Systems that decide, explain, and hold up in production.

Ads Systems Instacart Work Selected Projects Resume

Production ML

Data, features, models, serving, experiments, rollout, and monitoring.

Decision Science

PhD Economics, UCLA. BA Statistics & Economics, UC Berkeley. Causal thinking applied in production.

Builder Taste

Products that must decide: what to bid, what to remember, whether to act.

At Instacart

Ads, availability, and marketplace decisions at Instacart.

My Instacart work spans ads bidding, ranking, causal marketing measurement, availability prediction, fulfillment, and production ML. A score is not enough. The system has to decide, run reliably, and measure what happened.

What I actually do

I own the full loop: problem framing, data, features, model and control design, serving, experiments, rollout, monitoring, debugging, and migration.

Built Instacart's first optimized ads automated bidding system from scratch. It replaced manual product-keyword bids with weekly-budget bidding across roughly 70% of Sponsored Products spend.
Started from ROAS equalization: estimate product and keyword returns, choose advertiser target ROAS, pace budgets, simulate bids, write bids, observe outcomes.
Built causal marketing measurement and in-app decision systems where observed conversions were not clean labels.
Built availability and fulfillment systems across noisy inventory signals, real-time LightGBM serving, calibration, rollout, and monitoring.

Optimized ads bidding

Weekly-budget optimization, ROAS estimation, pacing, bid simulation, and bidding loops for a real ads marketplace.

Causal and marketplace measurement

Coupon and in-app marketing systems where incrementality and selection bias matter.

Availability and fulfillment ML

Models and serving paths for product-store availability, forecasting, calibration, and operational signals.

Selected Projects

Projects that show taste.

These projects share a pattern: turn fuzzy intent into systems that decide, preserve trust, and show their work.

Live product Neighborhood software

Bountiful Garden

Bountiful Garden helps neighbors share backyard produce before it goes to waste. The goal is simple: make a local exchange happen.

It is founder-shaped work: product design, trust, and marketplace mechanics for ordinary neighborhood life.

Marketplace Community Full-stack product

Visit Bountiful Garden

Live product Agent memory v0.2.16

OpenClawBrain

Evidence, not vibes accountable local agent memory

Distill Corrections, preferences, workflows, and context.

Retrieve SQLite FTS finds the few memories that match this turn.

Inject A bounded XML block keeps the prompt small.

Prove Status, health, search, graph, proof, and route events stay inspectable.

LLM suggests meaning. Code enforces scope. SQLite stores the graph and evidence.

OpenClawBrain gives OpenClaw agents local, inspectable memory. It remembers durable corrections like “use pnpm here” without dumping raw chat history into every prompt.

The spine is simple: SQLite stores the graph, local LLMs propose memories, validators enforce trust boundaries, and proof pages show what was captured or injected. Current public release: 0.2.16.

See OpenClawBrain → How it works Verify it

Private R&D Autonomous trading

Project Pelican

Project Pelican is a private autonomous options research system. It joins data, forecasting, risk controls, execution, and monitoring.

It is still internal. It reflects the same interest: systems that keep making measured decisions under uncertainty.

Quant finance ML pipelines Autonomous systems

Pelican Overview

Patents

Applied ML systems, backed by patents.

My Instacart work spans automated bidding, marketplace decisions, causal targeting, retailer intelligence, and fulfillment optimization.

US 12,511,677 B2 — Automated Policy Function Adjustment Using Reinforcement Learning Algorithm

Issued Dec. 30, 2025. Covers reinforcement-learning control for automated bidding in content campaigns. The practical work translated advertiser objectives and outcomes into bidding policy updates for a real ads marketplace. Assignee: Maplebear Inc.; inventors include Jonathan Gu.

Issued patent Reinforcement learning Automated bidding

Broader patent portfolio

Additional filed or submitted work covers double-wide ad auctions, causal targeting, dynamic offers, retailer classification with LLMs, fulfillment optimization, replacement agents, and item display decisions.

Writing

Selected writing.

Writing is where I make the logic visible: what shipped, what failed, and what still seems true after the numbers come in.

Get in touch.

I am interested in applied ML, agent infrastructure, marketplace design, local software, and systems that make real decisions.

Reach me directly

Email is best for thoughtful notes about projects, collaboration, or roles that fit.

Email LinkedIn Twitter / X GitHub Resume (PDF)