Machine Learning Engineer 2

What You’ll Do

Build systems that ingest unstructured and inconsistent source material, such as policies, protocols, rules, and spreadsheets, and turn it into structured, versioned, and verified knowledge.
Build extraction, normalization, and validation pipelines using LLMs and classical techniques; detect conflicts, gaps, and stale information, and route uncertain cases to human review.
Build and optimize Retrieval-Augmented Generation (RAG) pipelines that give downstream systems and agents accurate, grounded, and traceable answers with low latency.
Build, ship, and maintain components of the agentic systems that act on this knowledge, including orchestration, tool calling, and state management.
Build evaluation for the systems you ship: ground-truth datasets, accuracy metrics, regression tests that run before every release, LLM-as-judge and human review, and monitoring of quality in production.
Diagnose and improve accuracy, latency, reliability, and cost of ML systems in production.
Build components that can be reused across DexCare’s AI products, and contribute to other products as priorities shift.
Follow ML lifecycle best practices: versioning of models, prompts, datasets, and knowledge; deployment; monitoring; and retraining.
Handle patient and customer data in line with HIPAA and DexCare’s privacy and security standards.

What You’ll Bring

3-5 years of experience in machine learning or software engineering, including building and running production ML systems.
Experience building systems that extract, structure, and verify knowledge from unstructured sources, such as documents, policies, or operational data, with measurable accuracy.
Hands-on experience shipping LLM-based agentic systems to production, including orchestration, tool calling, and handling failures with retries and fallbacks.
Experience building retrieval systems and improving their accuracy and latency.
Experience evaluating ML and LLM systems: building eval datasets, defining metrics, and using the results to decide what ships.
Strong Python and software engineering fundamentals, including data structures, algorithms, testing, code review, and API design.
Familiarity with MLOps practices and at least one major cloud platform.
Clear communication and the ability to work across product, engineering, and data teams, including across time zones.
Comfort working in ambiguous problem spaces where requirements and data evolve.

Preferred

Experience with agent frameworks such as LangGraph or Google ADK.
Experience with Azure / AWS and their ML and AI services.
Experience with knowledge graphs, rules engines, or other structured knowledge representations.
Experience with predictive ML or optimization for problems such as forecasting, scheduling, ranking, or matching.
Experience with guardrails, safety classifiers, or red-teaming for LLM applications.
Experience in healthcare or another regulated industry, including handling PHI under HIPAA.