Head of Machine Learning

San Franciscofull-timeexecutive

Machine LearningData & AI

0 views0 saves0 applied

Apply Now

Quick Summary

Overview

Position Overview Bobyard solves hard ML problems to automate construction. Contractors waste dozens of hours per project on manual takeoffs. We replace that with model systems that have to be better than human — on messy, high-stakes data.

Key Responsibilities

Build SOTA deep learning systems: Architect models that read, reason, and predict across vision, language, and structured data. One error costs customers millions. 90% isn’t good enough.

Technical Tools

deep-learningetlmachine-learning

Bobyard solves hard ML problems to automate construction. Contractors waste dozens of hours per project on manual takeoffs. We replace that with model systems that have to be better than human — on messy, high-stakes data.

We’ve built a high-performing ML team taking on the toughest problems in applied AI. Now we need a leader to set technical direction, kill distractions, and 10x execution speed and quality.

This isn’t a “strategy only” role. The bar is: you’re as technical as our best ICs. You’ll define the roadmap, grow the team, and hold a savage hiring bar. You’re in the code, the papers, and the prod metrics — because that’s where alpha comes from.

Responsibilities

~2 min read

→
Build SOTA deep learning systems: Architect models that read, reason, and predict across vision, language, and structured data. One error costs customers millions. 90% isn’t good enough.
→
Own end-to-end ML: Data pipelines, labeling, training, eval, and low-latency inference. If it doesn’t ship to prod at scale, it doesn’t matter.
→
Scale the infra: Train on massive datasets and serve millions of predictions. Push accuracy up and cost down. Make the hard systems calls.
→
Guarantee reliability: Uptime, latency, drift, degradation. When models fail, revenue stops. You own every prediction.
→
Push performance: Accuracy, robustness, generalization. Real-world data is ugly — scanned, incomplete, adversarial. Make it work anyway.
→
Ship with product + engineering: No research islands. You tie every model to user impact and revenue. Integrate into features customers use daily.
→
Hunt failure modes: Live in production data. Find where we break, why we break, and fix it this sprint.
→
Apply research pragmatically: Stay at the frontier of deep learning. Take what works from papers, ship it next week, measure impact.
→
Set technical strategy: Define the ML roadmap. Decide what we build in-house, what we leverage, what we kill. Force focus on the highest-leverage bets.
→
Lead and grow the team: Hire elite ML talent. Coach, unblock, and develop world-class engineers. Give direct feedback. Keep the bar brutal.

Founder-level ownership: You treat model quality like it’s your own P&L. You see problems before anyone else and you’ve already fixed them.
Deep learning authority: You’ve trained and shipped large-scale DL systems — CV, NLP, multimodal, or beyond. Transformers, diffusion, whatever it takes.
Production-obsessed: You’ve put ML into products with real SLAs. You understand the gap between demo and 99.9% reliable.
Messy data killer: Real-world data is garbage. You’ve turned chaos into training sets that produce models customers bet their business on.
Research-minded, execution-obsessed: You read arXiv, but you ship. You optimize for impact, not papers.
Proven technical leader: You’ve led a world-class ML team. Top engineers want to work for you because you make them better and you ship.
Startup work ethic: This isn’t a cushy big-tech director role. We’re at war with manual workflows in a $13T industry. You want that.
Learning machine: New architecture drops, you’re fine-tuning it same day. You make unknown spaces tractable for the team.

This is full-time & in-person in SF. Learning rate and dedication beat pedigree. If you can prove you’ll lead us to ship models customers trust with million-dollar decisions — at the speed and quality the market demands — we want you.