ML Ops Engineer

Solana; Chicago (IL); On-site; USD 120,000 - 180,000; Full time

Job summary

Coinflow, a leading PSP focused on real-time payments and fraud prevention, seeks an MLOps Engineer to own the end-to-end model ops stack. You will build and operate the inference layer for real-time scoring in the authorization path and deploy robust ML pipelines with point-in-time accuracy.

You’ll collaborate on model deployment strategies, drift monitoring, and integration with external risk vendors. This role combines hands-on ML infrastructure with safety-first controls and scalable

Qualifications

• 4+ years in MLOps, ML infrastructure, or ML platform engineering in fintech/payments or regulated environments.
• Experience deploying models with real-time latency requirements in production.
• Strong Python skills; comfortable with TypeScript/Node in services.
• Hands-on with ML pipeline/tools and judgement on tool usage.
• Solid infrastructure skills: containers, Kubernetes, CI/CD, and cloud providers.
• Experience monitoring models in production and linking metrics to business outcomes.
• Familiarity with fraud, risk, or credit modeling and explaining model behavior to non-ML stakeholders.
• Comfort with AI-native tooling as daily driver or eagerness to start.

Responsibilities

• Real-Time Model Serving: build/operate inference layer in authorization path with low latency and safe fallbacks.
• ML Pipelines & Feature Store: own training, validation, deployment pipelines end-to-end with point-in-time correctness.
• Model Deployment & Rollout: manage versioning, registries, canary rollouts, and one-click rollback with eval-driven decisions.
• Monitoring & Drift Detection: track latency, score distributions, drift, approvals, fraud outcomes; integrate alerts.
• Third-Party Model Integration: evaluate external vendors, run shadow testing, compare performance on our data.
• Merchant-Facing Risk Tooling: power data and scoring behind merchant risk products.

Skills

Python

TypeScript

ML pipelines

Kubernetes

Cloud platforms

Production engineering

Tools

MLflow

Kubeflow

SageMaker

Vertex AI

BentoML

Feast

Job description

About Coinflow

Coinflow is the next-generation payment service provider revolutionizing global financial infrastructure with stablecoins, AI-driven fraud prevention, and instant settlement. Coinflow enables businesses to grow faster with instant settlement, fraud & chargeback indemnity, global pay-ins, multi-currency FX, and unified payouts. Founded in 2023, the company serves marketplaces, fintechs, remittance providers, gaming platforms, and ecommerce merchants worldwide.

Since our seed round in 2024, we've achieved 23x revenue growth and scaled to multi-billion-dollar annual transaction volume. In response to this growth, Coinflow announced a $25M Series A in October 2025 - led by Pantera Capital, CMT Digital, Coinbase Ventures, Jump Crypto, and Reciprocal Ventures - accelerating our mission to power the world’s fastest-moving businesses with innovative, reliable global payments.

Coinflow is proudly headquartered in Chicago, IL. Learn more at coinflow.cash.

About The Role

We're hiring an MLOps Engineer to own the platform that takes our fraud, risk, and payments models from notebook to production. You'll build the pipelines, serving infrastructure, and monitoring that let us score every transaction in real time, retrain safely.

Our models sit in the authorization path of real money, so latency, reliability, and explainability matter as much as accuracy. You'll be hands-on with AI-native tooling: we use AI as a force multiplier for building pipelines, writing evals, and automating the boring parts of model operations. If you want to be one of the first ML infrastructure engineers building this way at a fast-growing PSP, you'll fit in well here.

What You'll Own

Real-Time Model Serving: Build and operate the inference layer that scores transactions in the authorization path. Hit tight latency budgets, design safe fallbacks when a model or feature is unavailable, and make sure a model outage never becomes a payments outage.

ML Pipelines & Feature Store: Own training, validation, and deployment pipelines end to end. Build reproducible feature pipelines from our transaction, device, and chargeback data, with point-in-time correctness so offline training matches online scoring.

Model Deployment & Rollout: Make shipping a model boring. Versioning, registries, shadow mode, champion/challenger, canary rollouts, and one-click rollback. Every change is gated by evals, not vibes.

Monitoring & Drift Detection: Instrument models the way we instrument services. Track latency, score distributions, feature drift, approval rates, and fraud and chargeback outcomes, wired into our observability stack with alerts that fire on real problems.

Third-Party Model Integration: Integrate and evaluate external fraud and risk vendors alongside our own models. Run parallel and shadow testing, compare performance on our data, and give the team a clear answer on what to trust.

Merchant-Facing Risk Tooling: Power the data and scoring behind merchant-facing risk products, so merchants can see risky customers and set their own blocking rules on top of our signals.

What We're Looking For

• 4+ years in MLOps, ML infrastructure, or ML platform engineering, ideally at a fintech, payments company, or other regulated environment
• You've run models in production with real latency and uptime requirements, not just trained them. Experience with real-time inference in a transaction or decisioning path is a big plus
• Strong Python and solid production engineering habits. Comfortable reading and writing TypeScript/Node, since that's where most of our services live
• Hands-on experience with ML pipeline and serving tooling (MLflow, Kubeflow, SageMaker, Vertex, BentoML, Feast, or similar) and the judgment to know when not to use them
• Solid infrastructure skills: containers, Kubernetes, CI/CD, and at least one major cloud provider
• Data engineering chops: you can build reliable feature pipelines over operational, warehouse, and streaming data, and you understand training/serving skew
• Experience monitoring models in production: drift, calibration, and tying model metrics back to business outcomes like fraud loss and approval rates
• Familiarity with fraud, risk, or credit modeling, and with explaining model behavior to non-ML stakeholders like banks and auditors
• Comfort working with AI-native tooling (Claude Code or similar) as a daily driver, or genuine excitement to start
• A bias toward shipping. We'd rather have a working v1 of a model pipeline today than a perfect platform next quarter

The base salary range for this role is $120,000 - $180,000 USD + equity. The actual base salary offered depends on a variety of factors, including but not limited to experience, education, skills, qualifications, and business needs.

Join the team rewriting how money moves worldwide—and become a driving force in the $194 trillion cross-border payments market.