Senior Machine Learning Engineer

Senior Applied ML Engineer, Marketing Systems About the role We’re building a product that helps teams plan, launch, and improve marketing outcomes by combining automated decisioning with robust data + platform integrations . We’re looking for a Senior ML Engineer who is equally comfortable building applied ML systems and the backend infrastructure that makes those models reliable in production. You’ll work on the full lifecycle: turning messy marketing data into usable signals, building models that make better decisions over time, and wiring everything into a platform that can execute safely, measure results, and learn continuously. Compensation : Founding Level Equity + Salary What you’ll do Applied ML / Decisioning Build systems that recommend and optimize marketing actions (e.g., budget allocation, pacing, creative/ad selection, audience targeting signals, channel mix). Develop learning loops from outcomes: experiment design, counterfactual/holdout analysis (when applicable), offline evaluation, and online monitoring. Implement ranking/bandits or constrained optimization approaches where they fit (ROI under guardrails, budget constraints, frequency caps, etc.). Work with noisy, delayed feedback signals (attribution limitations, conversion lag, partial observability). Data + Measurement Design data models for campaigns, spend, conversions, events, and identity signals; unify across platforms. Build pipelines for ingestion, normalization, deduping, and reconciliation (spend/conversion mismatches, late-arriving data, API quirks). Improve measurement robustness (server-side events, event schemas, model features, privacy-aware aggregation). Backend / Platform Engineering Build backend services/APIs that expose decisioning outputs and integrate with execution workflows. Implement orchestration primitives: queues, schedulers, state machines for “plan → launch → monitor → adjust.” Engineer for production realities: rate limits, retries, idempotency, backfills, observability, and SLAs. Create internal tooling for debugging decisions (why the system did X), data QA, and replay. What we’re looking for 5–10+ years of experience shipping production systems, with meaningful time in applied ML . Strong coding ability and software fundamentals (you can build real services, not just notebooks). Experience with at least some of: Applied ML for optimization, ranking, forecasting, or decisioning under constraints. Marketing/adtech data (campaign hierarchy, spend, ROAS, CPA, conversion lag, attribution caveats). Production ML: feature pipelines, model training/evaluation, deployment, monitoring, model/data drift. Building backend systems that handle messy external APIs and high data volume. Comfortable working in ambiguity and iterating quickly with product/customer feedback. Nice to have Hands‑on experience integrating or working with major ad APIs (Meta/Google/TikTok/etc.). Experience with experimentation platforms, bandits, uplift modeling, or causal inference in real systems. Familiarity with privacy/measurement shifts (CAPI, iOS changes, consent modes). Experience designing guardrails for automated systems (budget caps, safety checks, “do no harm” constraints). Example projects you might work on A decision engine that recommends budget shifts daily under pacing + spend constraints. A “next best action” system for creative/ad set selection based on recent performance + fatigue signals. A cross‑channel performance model that normalizes platform‑specific metrics into comparable signals. A robust ingestion + reconciliation pipeline that can backfill and explain discrepancies. A service that produces explainable recommendations with audit logs (“what changed, when, and why”). Tech stack ML: Python, PyTorch/TF, scikit‑learn, XGBoost; orchestration (Airflow/Dagster); feature store optional Backend: Python/Node/Go; REST/gRPC; queues (Kafka/SQS/PubSub) Data: Postgres + warehouse (BigQuery/Snowflake); Redis Infra: AWS/GCP; Docker/Kubernetes; Terraform Observability: Datadog/Grafana/Sentry #J-18808-Ljbffr