ML / Optimization Engineer

About aematic

aematic is the optimization and measurement layer for advertising on AI answer-engines, starting with OpenAI’s ChatGPT. As advertising opens up on AI assistants, advertisers need real tooling to plan, launch, optimize, and measure campaigns on a channel the native consoles barely support. aematic runs the full loop for the advertiser: research the account, plan and launch campaigns, manage bids, and measure outcomes. The optimizer is the product, and you build the engine inside it.

The role

You own the algorithms that decide how to spend an advertiser’s budget. That spans automated bidding against target-CPA goals, budget allocation, creative and angle testing, and the agentic loop that plans and adjusts campaigns. Our founding team has built and run ad-bidding systems at scale, and you are building the equivalent intelligence for the AI ad channel from the ground up. This is hands-on applied ML and optimization, not research for its own sake; the work ships and moves real spend.

What you will do

• Design and build the bidding and budget-optimization systems that manage live ad spend toward advertiser profit.
• Develop and improve the multi-agent loop (campaign planning, persona and angle research, optimization) that runs on top of large language models.
• Build experimentation and learning systems so the optimizer improves as it manages more spend across more advertisers.
• Turn sparse, noisy, fast-moving auction data into reliable bidding decisions.
• Work with data scientists on measurement and with backend engineers to put models into production.

What we are looking for

• 4+ years in applied ML, optimization, or quantitative engineering, with strong Python.
• Experience taking models or algorithms into production, not just notebooks.
• Background in one or more of: bidding, recommendations, pricing, marketplaces, reinforcement learning, or other sequential decision systems.
• Comfort working with uncertainty: small data, shifting auctions, and incomplete feedback.
• Experience building with LLMs and agentic systems, or strong motivation to.

Nice to have

• Direct experience with ad auctions, real-time bidding, or budget pacing.
• Familiarity with causal inference or incrementality measurement.

Why now

The bidding intelligence you build is the core of the product, and the channel it runs on is brand new. This is a rare chance to build a category-defining optimization engine from scratch.