Role overview
This position is for an ML engineer within the ML Platform team, dedicated to teams working with large scale market data.
You will be the primary ML Platform contact for a functional team at CFM, dedicating a majority of your time to enabling that team’s ML work, and the remaining time to collaborating with the core ML Platform group.
The goal of the role is to make the default ML workflow faster and safer for large scale market data users by:
building and improving Python-first tooling and patterns,
ensuring solutions are production-ready (MLOps, reliability, monitoring),
and occasionally diving into C++ parts of the stack to debug issues, investigate performance bottlenecks, or contribute fixes in collaboration with owners.
This is an enablement role: success is measured by team productivity, fewer recurring failures, and adoption of shared patterns, not by isolated heroics.
What you'll work on
Enable and accelerate a functional team working with full scale market data by supporting their end-to-end ML lifecycle (data training evaluation
- deployment).
- Drive adoption of ML Platform tools and services through hands-on integration support, examples, and pragmatic guidance.
- Guide the evolution of ML Platform tooling based on real user needs (identify friction, propose improvements, validate with users, help ship changes).
- Establish and promote standards for ML development: reproducibility, quality, auditability, and maintainability (testing, versioning, documentation).
- Build self-service tooling (libraries, templates, reference implementations, automation) to reduce dependency on the platform team.
- Improve production readiness of ML systems: CI/CD, environment consistency, monitoring/alerting, incident readiness, and safe rollout practices.
- Mentor junior team members as the team expands; teach by building (docs, examples, office hours, paired debugging).
- Advocate for industry best practices in ML-related software engineering across the company.
What we're looking for
Technical:
Experience as a Data Scientist (useful for empathy with research workflows and evaluation practices).
Experience with inference servers (e.g., Triton) or building production model-serving services (HTTP/gRPC, scaling, latency/throughput tradeoffs).
Platform design / software architecture experience (APIs, multi-tenant systems, shared libraries, backwards compatibility).
Experience with C++/Python interoperability (e.g., bindings) and performance profiling across language boundaries.
“Design thinking” applied to platform work: identifying user journeys, reducing cognitive load, making the right thing the easy thing.
If you don’t meet every requirement but believe you’d be a great fit, feel free to reach out to us.