Full-stack Data + ML Engineer (Mid-Level)

Full-stack Data + ML Engineer (Mid-Level)

Role & responsibilities:

Data Engineering (50%)

• Design and build robust

  ELT/ETL pipelines

  from app, events, and 3rd-party data sources (batch & streaming).
• Create well-modeled data layers (staging/marts) with

  testing, documentation

  , and version control (e.g., dbt).
• Operate and optimize

  data warehouses/lakes

  , ensuring data lineage, quality checks, and secure access (PII compliance).
• Contribute to

  observability, cost tracking,

  and on-call support for data pipelines.

ML/AI (50%)

• Frame business problems, prepare datasets, and

  train/evaluate ML models

  for production use.
• Build and maintain

  inference services/APIs

  (e.g., FastAPI, Triton, KServe) with defined

  latency and cost targets

  .
• Implement

  LLM pipelines (RAG)

  , manage retrieval evaluation, prompt optimization, and safety guardrails.
• Work on

  classic ML

  use cases such as

  risk scoring, recommendation, churn, uplift modeling

  , and A/B testing.
• Monitor model

  drift, data integrity,

  and performance; maintain detailed

  runbooks

  and documentation.

Preferred candidate profile

• 4 - 6 years delivering production data systems and ML features.
• Strong SQL + Python; hands-on with dbt and an orchestrator (Airflow/Prefect/Dagster).
• Experience with a cloud DW (Snowflake/BigQuery/Redshift) and lake formats (Parquet/Delta/Iceberg).
• ML toolchain: PyTorch/TF, scikit-learn, experiment tracking (MLflow/W&B).
• Model serving know-how; comfort with Docker, CI/CD, and basic Kubernetes concepts.
• Clear communication, documentation habits, and pragmatic trade-offs.

Nice to have

• Streaming (Kafka/Flink/Spark Structured Streaming) or CDC (Debezium).
• Feature stores (Feast) and vector DBs (pgvector/FAISS/Weaviate) for LLM/RAG.
• Fintech exposure: lending/underwriting, bureau/alt-data ingestion, model risk controls, data retention & compliance.