AI Engineer, Intelligence & Optimization

Autonomous Ad-Optimization Agents • San Francisco

Location

San Francisco, California (on-site, full-time)

Compensation

$200k–$300k + equity · on-site in San Francisco · owns the decision layer

About the company

Our client is a fast-growing, Series-A-stage AI startup building autonomous agents that run paid digital advertising for consumer and tech companies. The agent runs paid acquisition across the major social and search ad platforms for mobile, gaming, AI and tech customers, replacing a human media-buying team end to end. Founded in 2023, they have raised $10M to date and are raising a Series A at a $100M valuation, on roughly $30M of revenue and about 30% month-over-month growth. It is a flat organisation of around twenty people across four offices on three continents, headquartered in San Francisco.

The role

You would own the intelligence and decision layer specifically — a separate platform team owns the simulator and the tooling underneath it. The scope is the recommendation and scoring systems behind bid changes, budget reallocation, pause and boost calls, and postback optimisation; designing the learning loop so the strategy improves over time instead of resetting; writing optimisation policies over noisy live data, from rule-based through post-training and reinforcement learning; and orchestrating how multiple agents reason over campaign context. It is a flat, high-autonomy team with no dedicated product layer, so this suits someone who ships without waiting to be directed. You would report to the CTO and co-founder, who reviews every candidate profile himself. They are hiring two people for this role.

What you'll do

Own the decision layer: recommendation and scoring systems for bid changes, budget reallocation, and pause/boost calls
Design the learning loop so campaign strategy compounds over time rather than resetting
Write optimisation policies over noisy live data — rule-based, post-training and reinforcement learning
Orchestrate how multiple agents reason over campaign context
Design experiments and reason statistically about results under real financial stakes

What we're looking for

These are hard requirements. Check each one that applies to you.

I have shipped a closed-loop decision system in production — logged outcomes feed back and change what the system decides next, as an ongoing loop rather than a one-off model
I have applied reinforcement learning, bandit algorithms or policy optimisation to a live production system, not only in research or coursework
I hold a quantitative STEM degree — engineering, physics, mathematics, computer science or statistics
I can work on-site in San Francisco full-time
I already hold transferable US work authorisation — for example an active H1B eligible for transfer, or OPT/CPT — or I am a US citizen or Green Card holder; I understand this client will not file a new, first-time sponsorship
I have 3 or more years shipping production machine-learning or optimisation systems
I am self-directed and ship without waiting for direction

Please note: every item above is a hard requirement for this role. Applications that don't meet all of them will be rejected. We recommend applying only if you check every box.

Don't check every box? No worries — submit your info here and we'll keep you in mind for roles that could be a better fit.

Submit Your Info

Nice to have

Direct experience with ads optimisation mechanics — bid pacing, budget allocation under spend caps, ROAS or CPA targets, postback optimisation
A background at, or selling into, companies running paid acquisition at scale across the major ad platforms
Recommendation or scoring systems — not just ranking — that moved a real business metric rather than an offline benchmark
Comfort with statistical reasoning under noisy production data, including experiment design

Tech stack

Python, machine learning, recommendation systems, reinforcement learning, ad optimisation and bid management systems.

Compensation & logistics

$200k–$300k base + competitive equity
On-site in San Francisco, published by the client as full-time in the office
Open to visa transfers only — for example OPT or an H1B transfer. The client will not file a new, first-time visa sponsorship
3+ years shipping production machine-learning or optimisation systems
2 hires planned
Three interview stages
Full-time

Back to all roles