The company builds finance infrastructure that sits inside every payment a business makes.
The work is high-stakes, dense with data and leaves little room for error.
The team values drive and urgency, and it weighs what you have built above where you studied. Everyone owns problems end to end and makes decisions that change the result.
Median customers save 7.5% and grow revenue 16% in their first year, well ahead of businesses without the platform.
Most agent work today runs one agent on one trace, with a person reading the output. This role targets systems where that breaks down: thousands of agents searching a large space in parallel, each returning partial results, with an orchestrator combining them into something usable, such as hundreds of merged pull requests, a dataset, an input for another system, or something not yet built.
You will design and run these systems, covering how work is split, how agents coordinate, how results are deduplicated, verified and ranked, and what the whole thing costs. The problems include agent swarms, agentic map-reduce and long-running fan-out workflows, and you take them from prototype to production.
Responsibilities
Decide how to break a problem apart, what each agent sees, how many agents to run and which model handles each step.
Raise accuracy through context design, self-verification, and voting or consensus across agents, and recognize when adding agents stops helping.
Assess quality without reading every trace, using sampling, automated checks, and precision and recall on the final output.
Lower cost and latency per useful result through model routing, caching, early stopping, and smaller or fine-tuned models where they perform well enough.
Convert one-off experiments into repeatable workflows on the existing orchestration stack.
Test new models, inference setups and training approaches, and adopt what works.
Requirements
You have shipped applied AI systems and LLM applications.
You have hands-on work with multi-agent or high-fan-out LLM workloads, such as parallel agents, map-reduce over models, or large batch inference.
You write strong Python and build concurrent, distributed systems that run reliably in production.
You can judge output: you define what a good end result looks like before scaling, and you separate signal from agent noise.
You track cost, yield, false-positive rate and whether a change actually helped.
You can run a roadmap and a production system with little oversight.
You communicate clearly in writing.
Nice-to-Have: you have built agent orchestration or job systems, such as Ray, Temporal, Modal or custom ones.
Nice-to-Have: you have used LLMs to search large spaces like codebases, document sets, logs or other large corpora.
Nice-to-Have: you have post-trained or hosted open-source models for high-volume inference.
Nice-to-Have: you have a distributed systems background.
Nice-to-Have: you have public work we can review, such as an agent framework, an open-source project, a write-up or a paper.
Conditions
Benefits are listed separately for the United States, Canada and the United Kingdom.
Applicants with arrest or conviction records will be considered, in line with San Francisco's Fair Chance Ordinance.
Benefits
United States: flexible PTO.
United States: home-office equipment ordered through a central process.
United States: stipend for health and wellness spending.
United States: budget for travel between offices.
United States: weekly coffee stipend.
United States: 100% coverage of medical, dental and vision insurance for you, with partial coverage for dependents.
United States: annual One Medical membership.
United States: 401(k) with employer match on contributions made while you are employed by the company.
United States: fertility HRA of up to $10,000 per year.
United States: up to 16 weeks of birthing and bonding leave, or 8 weeks for bonding only, at 100% pay.
United States: pet insurance.
United States: in-office perks including lunch, snacks, drinks and more.
United States: relocation expense coverage to NYC or SF, if needed.
Canada: group medical, dental and vision coverage through Sun Life.
Canada: life, AD&D and disability coverage.
Canada: fertility drug coverage up to $4,000 lifetime.
Canada: group retirement plan with employer match, through RRSP and DPSP.
Canada: up to 16 weeks of birthing and bonding leave, or 8 weeks for bonding only, at 100% pay, with extra time available at reduced pay.
Canada: employee assistance program and virtual care through Lumino Health.
United Kingdom: private medical insurance through Freedom Elite.
United Kingdom: virtual GP and at-home care through eMed x Livi.
United Kingdom: workplace pension through Penfold, with a salary sacrifice option.
United Kingdom: up to 16 weeks of birthing and bonding leave, or 8 weeks for bonding only, at 100% pay, with extra time available at reduced pay.
Where you’d work
Fully remote
You can work from
United States
Relocation offered
About the company
Hidden companyCompany hidden
Industry: FinTech
Good
Your chances
Still hiring, not crowded yet, and you'd be among the first.
18 checks run
7 good signs
2 red flags
Still hiring?
11 checks
Actively hiring
StaleHiring
In its favour4
Still on the company's own careers site, checked 1 h agoModerate evidence
Found in the last 48 hours, before the big job boardsModerate evidence
Specific about the basics: pay, place, level, stack and contract all statedSlight evidence
1 moreFewer
States its salarySlight evidence
How crowded?
7 checks
Low
QuietCrowded
In its favour3
In its Early Window: not on the big job boards yetStrong evidence
Remote within United States onlySlight evidence
Senior level: far fewer people qualifySlight evidence
Against it2
Offers relocation: people from other countries apply tooSlight evidence
Pays more than 89% of similar roles: that draws applicantsSlight evidence
?
Fits Me
How well does this role fit you?
Answer a few questions or drop your CV, and every role gets a fit score with the reasons, this one first.