Note: The job is a remote job and is open to candidates in USA. Apt is seeking a Senior AI Engineer to build and operate production-grade AI capabilities for priority use cases. The role focuses on developing LLM applications, agents, retrieval-augmented generation workflows, tool and API integrations, evaluation systems, and reliable backend services while partnering with cross-functional stakeholders to deliver secure, observable, and maintainable AI products.
Responsibilities
- Design, build, test, and operate AI-enabled product capabilities across backend services, APIs, data flows, agent workflows, and user-facing experiences where appropriate
- Translate product requirements, customer jobs-to-be-done, clinical or operational workflows, and stakeholder feedback into concrete technical designs and working AI capabilities
- Take ownership of implementation quality from prototype through production, including maintainable code, documentation, testing, observability, issue resolution, and iterative improvement
- Partner with product and engineering teams to sequence work into shippable increments that deliver measurable value while managing technical risk
- Build LLM-powered applications, agents, agent components, and multi-step workflows that can use tools, retrieve context, follow defined instructions, and complete targeted tasks reliably
- Define and implement agent behavior, task flow, prompt structures, tool-use patterns, fallback behavior, escalation paths, and workflow control logic
- Develop reusable patterns for agent workflows, tool invocation, state handling, structured outputs, context windows, memory, and human-in-the-loop checkpoints where appropriate
- Diagnose agent failures and improve behavior through prompt refinement, retrieval tuning, workflow redesign, evaluation results, telemetry, and user feedback
- Design and implement retrieval-augmented generation workflows that ground AI outputs in trusted enterprise data, content, knowledge, and source systems
- Partner with data, ontology, knowledge, and engineering teams to define retrieval strategies, source attribution, indexing, chunking, metadata, ranking, traceability, and response-grounding expectations
- Build and tune context pipelines, vector or hybrid retrieval patterns, knowledge integrations, and reusable data-access patterns that support high-quality AI outputs
- Identify reusable retrieval, context, tool, API, and integration patterns that can be shared across multiple AI use cases and product teams
- Build and maintain evaluation assets such as test datasets, golden-answer sets, scoring rubrics, prompt/version comparisons, regression tests, and quality dashboards for AI-enabled capabilities
- Measure and improve agent and LLM application performance across accuracy, groundedness, task completion, latency, cost, safety, response quality, and workflow reliability
- Use logs, traces, telemetry, user feedback, evaluation harnesses, and production signals to identify failure patterns and improve system behavior over time
- Partner with Quality Engineering and Product teams to ensure AI capabilities meet acceptance criteria, responsible AI expectations, and release-readiness standards
- Integrate AI capabilities with approved APIs, enterprise systems, tools, services, data products, knowledge repositories, and platform capabilities using secure and governed patterns
- Build tool wrappers, action interfaces, integration utilities, and workflow components that allow AI systems to interact with business processes in controlled and auditable ways
- Collaborate with engineering, architecture, security, privacy, and operations teams to ensure integrations follow enterprise standards for authentication, authorization, monitoring, error handling, and supportability
- Design AI-enabled workflows with clear boundaries around what the system can retrieve, recommend, automate, escalate, or hand off to a human
- Apply strong software engineering practices to AI systems, including code quality, tests, CI/CD, deployment discipline, monitoring, runbooks, incident follow-up, and operational support
- Optimize AI capabilities for latency, cost, reliability, scalability, maintainability, traceability, and user experience
- Document architecture decisions, implementation patterns, evaluation approaches, known limitations, and operational handoffs so teams can maintain and extend AI capabilities responsibly
- Mentor peers and contribute to shared engineering standards, reusable patterns, code reviews, design reviews, and practical guidance for applied AI delivery
Skills
- Proven ability to ship production software with real reliability, latency, maintainability, and operational support expectations
- Hands-on experience building LLM-powered applications, agents, RAG workflows, tool integrations, evaluation harnesses, or other applied AI systems beyond simple demos or thin API wrappers
- Strong backend engineering fundamentals, including Python, APIs, cloud services, data flows, testing, CI/CD, observability, and production debugging
- Ability to reason through agent and AI system failures, including retrieval gaps, context problems, prompt failures, tool-call errors, hallucination risk, latency issues, cost challenges, and workflow breakdowns
- Strong product orientation and ability to partner with product, design, clinical, operational, and business stakeholders to ensure AI capabilities solve meaningful problems
- Ability to balance innovation speed with privacy, security, reliability, evaluation discipline, auditability, cost, and responsible AI expectations
- Comfort operating in a regulated or high-trust environment where customer trust, safety, and operational reliability are essential
- Bachelor's degree in Computer Science, Engineering, Data Science, Information Systems, Mathematics, or related technical field
- Equivalent combination of education and relevant technical experience may be considered
- 6+ years of production software engineering, applied AI engineering, ML engineering, data-intensive application development, backend platform engineering, or related technology experience
- Meaningful hands-on experience building LLM-powered applications, AI agents, RAG workflows, prompt/context systems, evaluation assets, AI-assisted workflows, or comparable applied AI capabilities
- Experience designing and deploying end-to-end technical capabilities from requirements and prototypes through production implementation, monitoring, and iteration
- Experience working with product, design, engineering, data, analytics, quality, security, privacy, and business stakeholders to ship reliable software or AI-enabled product features
- Strong programming skills, preferably Python, with backend engineering experience across services, APIs, integrations, data flows, and production application components
- Hands-on familiarity with LLM application development, including prompt engineering, RAG pipelines, embeddings, vector or hybrid retrieval, tool calling, structured outputs, and evaluation methods
- Familiarity with agentic systems and orchestration patterns, including multi-step workflows, state management, memory, routing, decision logic, human-in-the-loop controls, and emerging frameworks such as LangGraph, LangChain, LlamaIndex, CrewAI, Google ADK, OpenAI Agents SDK, or comparable tools
- Understanding of cloud and production engineering practices, including containers, CI/CD, monitoring, logging, deployment, security, privacy, latency, cost, reliability, and operational support
- Ability to work with data and integration patterns, including APIs, event flows, data products, enterprise systems, searchable knowledge sources, and scalable architectures
- Ability to test and evaluate AI-enabled systems, including response quality, groundedness, source attribution, task completion, hallucination risk, latency, cost, safety, and reliability
- Experience shipping production AI agents, copilots, assistants, conversational experiences, tool-using AI workflows, or AI-enabled automation products
- Experience with GraphRAG, knowledge graphs, ontology-backed retrieval, enterprise search, context graphs, content grounding, or semantic retrieval patterns
- Experience developing evaluation suites, golden datasets, scoring rubrics, prompt/version comparisons, regression tests, observability dashboards, or quality metrics for LLM or agent systems
- Experience integrating AI systems with enterprise APIs, FHIR/healthcare data, customer-facing applications, CRM/CDP platforms, messaging systems, workflow systems, or internal developer platforms
- Experience with model routing, model selection, inference optimization, fine-tuning, LoRA/PEFT, or adapting models to domain-specific workflows
- Prior experience in AI product teams, AI-native companies, big tech, digital health, or product-driven engineering environments
- Demonstrated ability to translate ambiguous AI capabilities into concrete, reliable, user-facing features with measurable outcomes
- Healthcare, life sciences, financial services, or other regulated-industry experience preferred but not required
Company Overview