TixelJobs
B
Bask Healthvia muse

Applied ML Engineer, Agents & Automation

REMOTEPosted 6h ago
ML EngineerMid LevelFull-time#applied-machine-learning-engineer#machine-learning-engineer#ai-engineer#software-engineer#automation-engineer#remote

About the Role

About the Role We're looking for an Applied ML Engineer, Agents & Automation to design, build, and deploy intelligent systems that automate complex workflows across Bask Health 's telehealth platform. Bask Health brings together merchant operations, patient experiences, clinical workflows, pharmacy coordination, and commerce tools. Many of these processes involve multiple applications, structured information, business rules, and repetitive tasks. You'll help build AI-powered systems that can coordinate these steps while operating within clearly defined permissions and reliability requirements. This role focuses on developing production AI agents that use large language models, application tools, retrieval systems, and workflow orchestration to complete approved tasks. You'll work on the technical infrastructure that allows agents to retrieve information, select appropriate tools, execute actions, handle failures, and maintain context throughout multi-step processes. You'll combine applied machine learning with strong software engineering practices. Depending on the use case, this may involve model selection, fine-tuning, task-specific evaluation, or other machine learning techniques that improve agent performance. You'll also recognize when deterministic software and conventional automation are more reliable than model-driven decisions. Your work may support merchant operations, internal administrative processes, patient-related coordination, and pharmacy workflows. You'll collaborate with product managers, backend engineers, applied AI engineers, and operational stakeholders to identify opportunities where automation can reduce manual work and improve consistency. Because Bask operates in healthcare-related environments, agents must be designed with appropriate safeguards, access controls, auditability, and human approval requirements. You'll help ensure that automated systems operate within approved business and clinical boundaries and do not independently perform actions that require licensed professional judgment. This is a hands-on engineering position with meaningful ownership over agent behavior, automation architecture, and production reliability. You'll build reusable systems, measure their performance, and improve them based on real-world results. We're looking for someone who understands both the potential and limitations of AI agents and enjoys building systems that are useful, measurable, secure, and dependable. What You'll Do Design, build, and deploy AI agents that automate approved workflows across Bask Health 's platform. Develop agent architectures that combine large language models, tools, APIs, and application logic. Build multi-step automation systems that coordinate tasks across multiple services. Implement reliable tool-calling and function-execution capabilities. Develop agent orchestration logic for task planning, execution, and completion. Build reusable components for agent state management, context handling, and workflow execution. Integrate AI agents with internal APIs, databases, and approved external services. Develop retrieval capabilities that provide agents with relevant information and application context. Design structured outputs and validation systems for agent-generated actions. Implement permission checks and authorization controls for automated tool execution. Build human approval and confirmation steps for sensitive or consequential actions. Develop safeguards against prompt injection, unauthorized actions, and unreliable model behavior. Implement retry logic, error recovery, timeouts, and fallback mechanisms. Build monitoring and tracing systems that make agent behavior observable and debuggable. Create automated tests for tool selection, workflow execution, and task completion. Develop evaluation datasets and benchmarks for agent accuracy and reliability. Measure agent success rates, task completion, latency, and operational costs. Investigate agent failures and improve prompts, tool
Share