SavvyMoney - Lead ML/AI Platform Engineer

• 8+ years in software or ML engineering, including 5+ years shipping production ML systems and a track record of owning ambiguous, high-scope problems end to end. • Demonstrated technical leadership: you've shaped the ML strategy of a team or organization, mentored senior engineers, and been the person others rely on for difficult architectural calls. • Hands-on experience with both operating models we use: • AWS managed ML stack: Amazon SageMaker (training, tuning, hosted endpoints, model registry), Amazon Bedrock, and AgentCore for GenAI and agentic workflows. • AWS managed ML stack: • Open-source ML tooling: JupyterLab for notebooks, Spark for distributed processing, MLflow for experiment tracking and model registry. • Open-source ML tooling: • Deep working knowledge of the AWS stack — S3, Athena, Redshift, Glue, Step Functions, Lambda — plus SQL skills strong enough to model data for both analytical and ML workloads. • Production experience with GenAI/LLMs: RAG, prompt engineering, evaluation, and a clear grasp of the cost, latency, and safety trade-offs involved. • Familiarity with vector databases (e.g., pgvector, Pinecone) and sound judgment on when they're warranted versus alternatives such as NoSQL retrieval. • Working knowledge of Java sufficient to review service code, define API contracts, and debug integration issues with our microservices. • Deep expertise in Python and the core ML stack: scikit-learn, pandas, NumPy, PyTorch and/or TensorFlow, XGBoost / LightGBM. • Solid MLOps fundamentals — model monitoring, drift detection, reproducibility, experiment tracking, model registry, and cost observability — plus the ability to partner with DevOps on CI/CD rather than build it from scratch. • Excellent written and verbal communication — you can write both the design doc that aligns a dozen engineers and the one-pager that aligns the exec team. • Strong collaborator, comfortable operating in a role where scope is shared: you'll partner with a Data Scientist on models and DevOps on infrastructure, and you can navigate those seams while keeping clear ownership. • Ability to operate as an independent contractor through your own entity or an approved contracting arrangement, with reliable overlap with US Pacific business hours for architecture reviews and cross-team work. • Experience with ClickHouse or a comparable columnar / real-time analytical database. • Fine-tuning experience (LoRA / QLoRA, instruction tuning, or RLHF). • Streaming and real-time inference experience (Kafka, Kinesis, low-latency serving). • Infrastructure-as-code (Terraform, AWS CDK, CloudFormation). • Experience operating ML systems at meaningful scale — hundreds of millions of predictions per day, or equivalent. • Open-source contributions, conference talks, papers, or patents in ML / applied ML. • Who You'll Work With • Data Scientist — leads feature engineering, model training, fine-tuning, and experimentation. You'll partner on the research-to-production handoff, stand up the infrastructure they need, and jointly own the quality of what ships. • Data Scientist • DevOps — owns CI/CD, deployment infrastructure, and platform observability. You'll partner on ML-specific extensions (model artifacts, reproducible environments, canary and shadow deployments) rather than rebuilding what already exists. • DevOps • Backend engineers — own the Java microservices that consume ML outputs. You'll define the contracts and own the serving side. • Backend engineers • Product and engineering leadership — you're the technical voice on what we should build with AI/ML, and how. • Product and engineering leadership • Our Stack • Distributed processing: AWS Glue • Distributed processing: • Orchestration: AWS Lambda, AWS Step Functions • Orchestration: • Managed ML: Amazon SageMaker, Amazon Bedrock, AgentCore • Managed ML: • Open-source ML: JupyterLab, Spark, MLflow • Open-source ML: • Storage & query: Amazon S3 (data lake), Amazon Athena, Amazon Redshift, ClickHouse • Storage & query: • Languages: Python (ML, pipelines), Java (microservices)Why This Role Matters • Languages: