Careers at Mandate Labs
Open roles on the team building the issuer-side authorization layer for agent commerce.
About Mandate Labs
Mandate Labs builds the Decision Trust Protocol — real-time authorization infrastructure that evaluates the reasoning behind AI-agent transactions before money moves. As autonomous agents become participants in commerce, the card networks are building the identity and consent layer; we build the issuer-side decisioning layer that consumes those signals and renders the authorization decision. We are live in production with a paying card-program client, we publish our research openly, and we work at the intersection of payments infrastructure and frontier AI evaluation.
Applied ML Lead
San Francisco, CA (hybrid)
About the role
As the Applied ML Lead at Mandate Labs, you will own the Economic Decision Quality Score (EDQS) end to end: the scoring methodology, the production models, the public benchmark, and the dataset strategy. The research agenda is written and pre-registered in our published EDQS Research Framework; the production telemetry exists; this role is for the scientist-engineer who executes and extends both.
You will be our first dedicated ML hire, working directly with the founder. The evaluation problem — scoring the economic reasoning of autonomous agents under adversarial conditions — sits at the intersection of process supervision, chain-of-thought faithfulness research, and payments risk, and the data substrate is a production authorization path rather than a simulation.
Responsibilities:
- Execute the pre-registered EDQS benchmark on Vending-Bench 2: instrument runs with per-decision scoring, measure detection lead time and AUROC against transaction-fraud baselines, and publish results with code
- Run the adversarial evaluation program: red-team EDQS against score gaming, behavioral mimicry, trust-warming, and mandate splitting, and publish findings
- Own the production scoring stack — six anomaly detectors, behavioral drift detection over seven signal classes, confidence calibration — operating inside sub-100ms authorization budgets
- Define the schema, governance, and research use of our per-transaction reasoning and outcome dataset from live card programs
- Represent the research program publicly: arXiv submissions, workshop papers, and engagement with the agent-evaluation community
You may be a good fit if you have:
- Experience shipping ML to production in an adversarial domain — fraud, risk, abuse, or anomaly detection — where someone was actively trying to defeat your models
- Working fluency with the current agent-evaluation literature, including process supervision, reward hacking, and chain-of-thought faithfulness and monitorability
- Strong statistical rigor: pre-registered endpoints, calibrated uncertainty, and a preference for reproducible negative results over unreproducible positive ones
- Experience treating latency as a design constraint, with models that run inside production request budgets
- Comfort operating as an early hire with broad ownership and direct founder collaboration
Strong candidates may also have:
- A publication record or public benchmark and evaluation work
- Payments, fintech, or trust-and-safety domain experience
- Familiarity with behavioral biometrics or sequence anomaly detection
Representative projects:
- Demonstrate that EDQS degradation flags a Vending-Bench meltdown well before the bank balance collapses, and identify which dimensions carry the detection lead time
- Design the divergence metric between an agent’s attested reasoning and its observed behavioral fingerprint, robust to obfuscated traces
- Build the trust-warming detector that distinguishes a legitimately improving agent from an adversary farming the trust ladder
- Take the six anomaly detectors from heuristics to calibrated models without breaking the engine’s latency budget
The annual compensation range for this role is listed below. This role also includes equity.Annual Salary: $230,000 – $300,000 USD
Apply
Senior Backend Engineer
Remote — United States · Multiple positions · Open to exceptional candidates across the Americas
About the role
As a Senior Backend Engineer at Mandate Labs, you will build the authorization engine that card issuers depend on: the seven-gate decision pipeline, mandate and velocity enforcement, and the infrastructure that turns an agent’s request into an enriched decision inside a payment authorization window. The engineering bar is issuer-grade correctness — races, retries, and partial failures are design inputs in an authorization path.
You will work directly with the founder, a former Mastercard principal-member operator, on a production codebase that recently passed a comprehensive third-party security review.
Responsibilities:
- Drive sustained end-to-end latency to honestly measured targets at 150K+ authorizations per hour by collapsing decision-path database round-trips and extending the caching and reservation layer
- Own the correctness primitives issuer due diligence examines first: idempotency under parallel retries, atomic velocity reservation, and defined fail-closed/fail-open behavior
- Ship the v0.7 API: developer-experience endpoints, webhook event migration with backward-compatible aliases, and enum alignment with our published documentation
- Build the standards surface: ISO 8583 field mapping for issuer integration, network-compatible decision payloads, and the sandbox-to-production promotion flow behind our certification program
- Operate multi-tenant reliability: isolation, observability, and the SLOs our SLA commitments are measured against
You may be a good fit if you have:
- Significant senior backend experience with Python (FastAPI or similar), PostgreSQL, and Redis in concurrent, failure-sensitive systems
- Experience building or operating payments or financial infrastructure — authorization flows, issuer processing, ledgers, or comparable systems where money moves on your code’s decision
- A habit of benchmarking before claiming, with load tests run sustained rather than burst
- A track record of fixing classes of bugs rather than instances, with the postmortems to show for it
Strong candidates may also have:
- ISO 8583 or ISO 20022 exposure, or card-scheme certification experience
- Experience in SOC 2 or PCI-scoped environments
- SDK or developer-tooling work
The annual compensation range for this role is listed below and reflects the United States market. This role also includes equity. For candidates based outside the United States, compensation and conditions of hire are determined by candidate location, consistent with market standards for the role and region.Annual Salary: $160,000 – $200,000 USD
Apply
VP Operations
San Francisco, CA / Remote (US)
About the role
As VP Operations at Mandate Labs, you will build the operational function that converts enterprise interest into running card programs: the certification program that gates production access, the compliance posture that withstands bank due diligence, and the client operations that keep issuers renewing. This is a build-the-function executive role reporting to the CEO, with a natural path to COO as the company scales.
Responsibilities:
- Run the Mandate Certified Integration (MCI) program end to end for every new client — sandbox conformance, integration review, UAT sign-off, and production cutover
- Own enterprise onboarding, including client vendor-security reviews and InfoSec questionnaires, from signed agreement to live production
- Build compliance operations: the SOC 2 program, PCI scope management, third-party audit coordination, and the diligence evidence room
- Operate commercial machinery: order forms, pricing tiers and usage graduation in practice, SLA and status-page operations, and support escalation
- Manage the vendor and infrastructure-operations stack across deployment, monitoring, and compliance tooling
You may be a good fit if you have:
- Operations leadership experience in enterprise fintech or payments, including time on the receiving end of bank vendor due diligence
- Experience building — not only passing — certification or audit programs such as SOC 2, PCI-DSS, or card-scheme certifications
- Comfort operating across the United States and Latin America; Spanish proficiency is a plus
- The early-stage temperament to write the runbook and then run it yourself until there is someone to hand it to
Strong candidates may also have:
- Issuer-processor, banking-as-a-service, or program-manager operations experience
- Experience taking a company through its first enterprise client conversions
- Regulatory exposure in Latin American markets
The annual compensation range for this role is listed below. This role also includes equity.Annual Salary: $170,000 – $210,000 USD
Apply
Logistics
Minimum education: Bachelor’s degree or an equivalent combination of education, training, and/or experience.
Minimum years of experience: Years of experience required will correlate with the scope and level of the position.
Location-based hybrid policy: The Applied ML Lead role is based in San Francisco with hybrid flexibility. Backend engineering roles are remote-first within the United States, and we remain open to exceptional candidates elsewhere in the Americas. The VP Operations role is based in San Francisco or remote within the US, with travel to clients. All roles collaborate synchronously in US time zones.
Visa sponsorship: As an early-stage company, we evaluate visa situations case by case and will be transparent with you about what we can support before you invest time in our process.
We encourage you to apply even if you do not believe you meet every single qualification. Not all strong candidates will meet every qualification as listed. Research shows that people from underrepresented groups are more likely to doubt the strength of their candidacy, so we urge you not to exclude yourself prematurely and to submit an application if this work interests you.
Your safety matters to us. Mandate Labs recruiting communications come only from @mandatelabs.ai email addresses. We will never ask for money, fees, or banking information during the hiring process. If you are unsure about a communication, visit mandatelabs.ai/careers.html directly for confirmed openings.
How we’re different
We are a small team working on one problem: making autonomous agent transactions safe enough for the financial system to say yes to. That means we operate simultaneously as payments infrastructure engineers — with the correctness, latency, and audit obligations of an authorization path — and as applied researchers publishing openly on agent evaluation, behavioral telemetry, and decision quality. Our research is public, our pipeline documentation matches our code, and our claims are written so that a hostile reviewer can verify them. We believe the teams that win regulated infrastructure categories are the ones whose engineering culture survives due diligence.
Come work with us!
Mandate Labs is headquartered in San Francisco, with roots in Latin American payments infrastructure. We offer competitive compensation and equity, flexible time off, a remote-flexible culture centered on US time zones, and the kind of ownership that only exists when the category is being defined. Apply at [email protected] with the role in the subject line, and include a resume or LinkedIn profile.