Transform Healthcare Through AI‑Orchestrated Workflows
Build and operate Azure‑ready AI systems that power clinical workflows end‑to‑end. If you thrive at the intersection of AI engineering, orchestration, and product, and you can turn complex, multi‑model pipelines into reliable, cost‑efficient production systems, this role is for you.
About Zenara
Founder‑led US health‑tech startup building AI‑powered clinical software clinicians love. We combine cutting‑edge technology with deep healthcare expertise to solve real provider problems. Our culture values builders who ship, leaders who serve, and teams that deliver through clear communication and collaboration.
Why This Role Is Different
You architect and ship GA‑ready AI orchestration on Azure that directly improves patient care. You balance robust pipelines (prompt chaining, RAG, tool‑calling, multi‑model) with cross‑functional coordination so every feature is reliable, observable, low‑latency, and cost‑effective.
What Success Looks Like
- First 30 days :
- Ship improvements to deterministic multi‑step prompt chains for patient intake, validation, summarization, document generation, and outbound artifacts.
- Establish flows‑as‑code and prompt versioning with GitHub and environments.
- Provide clarity and hands‑on guidance on complex model / tool integrations.
- First 90 days :
- Operationalize multi‑model workflows (GPT‑4.1 / 5 + SLMs like LLaMA / Nous) with step‑specific model selection for latency / cost.
- Stand up automated evaluation workflows, telemetry, and dashboards (App Insights / OpenTelemetry).
- Implement resilience patterns : idempotency, retry / backoff, dead‑letter queues, rate‑limit handling.
- Azure GA stack in production for orchestration, scaling, and retries (e.g., Azure Functions Durable, Container Apps, Queues / Event Grid, KEDA).
- First 6 months :
- Act as the AI / technical hub for the team, mentoring engineers on prompt engineering, structured generation, tool‑calling, and agentic best practices.
- Align stakeholders; maintain clear documentation; drive sprints that deliver decisions and outcomes.
Core Responsibilities
Builder (40%)
Design and deploy LLM solutions for clinical workflows with deterministic, multi‑step prompt chaining.Implement structured generation pipelines and evaluation harnesses; build test suites for AI features.Operationalize RAG : ingestion pipelines from Azure Blob / ADLS Gen2, indexing in Azure AI Search (vector), chunking / embeddings, retrieval strategies.Integrate and debug complex multi‑model setups (GPT‑4.1 / 5 + SLMs) with step‑wise tool selection and cost controls.Standardize tool‑calling (MCP) for EHR / EMR, scheduling / communications, internal knowledge APIs; mix remote and custom‑hosted tools.Establish flows‑as‑code, prompt versioning, environments, and CI / CD (GitHub Actions) for controlled updates.Orchestrator (40%)
Translate AI / product vision into executable technical roadmaps; facilitate planning for AI releases.Operate a GA‑ready Azure stack for orchestration, retries, and auto‑scaling :Azure Functions Durable (stateful workflows), Azure Container Apps, Azure Queues / Event Grid, KEDA auto‑scaling.Observability with Azure Monitor / App Insights / OpenTelemetry; log / metric correlation per step / tool call.Reliability and performance :Interactive experiences : stream outputs and design responsive tool‑calls / RAG so users perceive fast feedback.Batch workflows : predictable end‑to‑end durations with clear SLAs and automated retries; timeouts aligned to flow needs.Concurrency : horizontally scalable execution with graceful handling under burst; autoscaling and backpressure controls.Cost : optimize architecture / model selection to meet budget goals; instrument and tune regularly for efficiency.Resilience : idempotency, retry / backoff, DLQs, circuit breakers, and rate‑limit handling across steps and tools.Communicator (20%)
Bridge technical and non‑technical stakeholders; run decisive meetings that produce decisions and unblock execution.Document processes, architectures, and technical decisions clearly; maintain Jira / Confluence hygiene and Git workflows.Promote psychological safety; surface hidden blockers early; facilitate AI‑focused problem‑solving.Technical Requirements
3+ years in LLM / prompt engineering, structured generation, evaluation workflows; daily user of AI assistants and automation tools.4–8 years hands‑on development with technical leadership / TPM experience; ability to break large projects into flows‑as‑code.Azure‑first stack :Orchestration : Azure Functions Durable, Azure Container Apps, Azure Queues / Event Grid, KEDA.Data : Azure Blob / ADLS Gen2, Azure AI Search (vector), CosmosDB / SQL, Redis.Observability : App Insights, Azure Monitor, OpenTelemetry.CI / CD : GitHub Actions; environments; PR‑based governance; prompt / version management.Web and services :TypeScript / React for front‑ends; Python / Node.js for services / workers / tooling.REST / gRPC; async jobs and streaming outputs; webhook / callback patterns.AI / ML :Multi‑model orchestration (GPT‑4.1 / 5 + SLMs like LLaMA / Nous) with step‑wise selection for latency / cost.RAG pipelines; embeddings; retrieval strategies; evaluation and guardrails.Tool‑calling / MCP; remote and custom‑hosted tool integration; EHR / EMR integrations preferred.Healthcare / regulatory :Familiarity with healthcare data models (FHIR) and regulated environments.Security / compliance : HIPAA, SOC 2 (HITRUST a plus); audit / readiness (Vanta or equivalent).Communication Requirements
Facilitate technical design sessions to consensus; translate between technical and business stakeholders.Run sprints / meetings that produce decisions; resolve conflicts constructively.Excellent written and verbal communication across cultures and time zones; strong async communication for remote work.Intangibles We Value
Proactive communication; crisp documentation of complex topics.Builds trust across technical / non‑technical teams; steady under pressure; bold when shipping.Sees conflicts and ambiguity as opportunities for better solutions.Major Advantages
Healthcare technology or regulated industry experience.Experience facilitating architecture reviews and technical decisions; flows‑as‑code governance.Stories of aligning misaligned teams; cross‑cultural collaboration; Azure AI stack familiarity (Foundry / Agent Framework awareness).What We Offer
Competitive compensation; 100% remote — work from anywhere.Direct mentorship from technical leadership; ideas ship if they work; minimal bureaucracy.Job Types & Benefits
Full‑time, Permanent, Contractual / Temporary.Internet reimbursement; Work from home.Schedule : Monday to Friday, US shift.Supplemental Pay : Joining bonus, performance bonus, yearly bonus.