Artificial Intelligence Engineer

Company Description

About TeleMEDix

TeleMEDix is building the next-generation clinical operating system—a full-stack, AI-powered platform that unifies clinical decision support, electronic health records, genomic intelligence, and value-based care analytics into a single intelligent layer. Our architecture is built on Azure Cosmos DB, large-scale biomedical knowledge graphs, NLP-driven clinical reasoning, and ontology-first design principles (HPO, MONDO, ICD-10, SNOMED CT). We are a pre-Series A healthcare AI company with deep domain expertise and a founding team that writes production code. Here some of our experience in AI:

• Healthcare AI, AI Agents development, and Clinical AI Decision Support Systems. Our national leadership and management experience and your proficiency with Azure, NLP, and FAISS. Highlight your hands-on experience with architecting healthcare knowledge graphs and biomedical NLP pipelines.
• Here is your opportunity to showcase your ability to deploy clinical AI systems on our national platform

The Role

We are looking for a Senior AI Engineer as a contractor at first who has significant hands-on experience architecting healthcare knowledge graphs, building biomedical NLP pipelines, and deploying clinical AI systems on Azure. This is not a research role—you will own and ship production infrastructure that clinicians depend on. You will work directly with the CEO/Architect and CTO to extend our 29-partition clinical Knowledge Graph, harden our ScispaCy NLP pipeline, and build the reasoning engine that powers our CDSS.

Key Responsibilities:

•  Knowledge Graph Architecture & Engineering

•      Design, build, and maintain a large-scale clinical Knowledge Graph on Azure Cosmos DB (Gremlin API) spanning 29 partitions across 26 clinical domains plus shared medication, guideline, and rule vertices.

•      Implement and validate KG seed pipelines (vertex/edge ingestion, Gremlin import files) from authoritative clinical guidelines (AHA, ACC, ADA, GOLD, KDIGO, etc.).

•      Architect the dual-track Clinical Rule Engine (CRE): Track 1 for structured threshold/stage/phenotype evaluation; Track 2 for guideline text and recommendation delivery.

•      Build and maintain HPO (17K+ terms), MONDO (25K+ terms), ICD-10, and SNOMED CT ontology infrastructure including cross-ontology bridging (MONDO↔HPO).

•      Ensure strict data separation: no patient data in the KG; patient data remains exclusively in InfEHR.

NLP & Clinical Language Processing

•      Own and extend the ScispaCy-based NLP pipeline for query Named Entity Recognition (NER) and InfEHR data normalization.

•      Ensure ScispaCy is always available as the primary NLP layer—LLMs are used only for human touch, formatting, fallback, and patient education; never for clinical decisions.

•      Build and tune entity linking, relation extraction, and negation detection models for clinical text (discharge summaries, lab results, problem lists, CCDA documents).

•      Integrate ontology-driven NER with the KG to enable deterministic, auditable clinical reasoning paths.

Vector Database & Retrieval Infrastructure

•      Architect and manage FAISS (or equivalent) vector indexes for semantic retrieval over clinical guidelines, drug references, and patient education materials.

•      Build and optimize RAG (Retrieval-Augmented Generation) pipelines that assemble clinical responses before LLM formatting—response assembly always occurs before LLM.

•      Implement embedding strategies for biomedical corpora using domain-specific models (BioBERT, PubMed, BERT, or equivalent).

Azure Cloud & Production Operations

•      Deploy and manage production services on Azure: Cosmos DB, Azure Blob Storage, Azure App Service / Functions, Azure Kubernetes Service (AKS) as needed.

•      Implement CI/CD pipelines, infrastructure-as-code, and monitoring/alerting for clinical AI services with zero-downtime deployment targets.

•      Manage cost optimization across Azure services; right-size Cosmos DB throughput (RU/s), storage tiers, and compute.

Clinical Domain & Quality

•      Collaborate with clinical leadership (SVP Clinical Standards) to translate evidence-based guidelines into computable KG structures and rule sets.

•      Build severity indexes alongside staging systems; implement Python-based risk stratification (not GNN).

•      Write and maintain comprehensive test suites; validate KG builds (target: zero-error ingestion across all clinical domains).

•      Maintain and contribute to ARCHITECTURE_MASTER.md as the single source of truth for all system design decisions.

Required Qualifications

•      5+ years of experience in AI/ML engineering with at least 3 years focused on healthcare or biomedical domains.

•      Proven track record architecting and deploying graph databases in production—strong preference for Azure Cosmos DB (Gremlin API), Neo4j, or equivalent.

•      Deep hands-on experience with biomedical NLP: ScispaCy, SpaCy, UMLS, MetaMap, or cTAKES. ScispaCy experience strongly preferred.

•      Expertise with clinical ontologies: HPO, MONDO, ICD-10, SNOMED CT, RxNorm, LOINC. Must understand ontology hierarchy traversal, cross-mapping, and phenotype-to-disease bridging.

•      Experience with any of these databases and embedding-based retrieval (MongoDB, Cosmos, FAISS, Pinecone, Weaviate, Milvus, or equivalent).

•      Production Azure experience: Cosmos DB, Blob Storage, App Service, Functions, AKS, or equivalent cloud-native services.

•      Expert-level Python; strong experience with FastAPI.

•      Solid understanding of clinical data standards: HL7 FHIR, CCDA/CDA, ADT, and healthcare interoperability (TEFCA/Carequality is a plus).

•      BS/MS in Computer Science, Biomedical Informatics, or a related field. PhD is a plus but not required—we value shipped production systems over publications.

Preferred Qualifications

Experience with any of the following will set you apart:

Pharmacogenomics / genomic data (e.g., DNA panels, PGx)

LangChain or equivalent RAG orchestration frameworks

Value-based care analytics, risk adjustment (HCC), ACO workflows

HIPAA/BAA compliance and healthcare security posture

.NET / C# (parallel codebase exists in .NET 8)

Revenue cycle management (RCM) and claims/EDI integration

React Native mobile development

Telehealth infrastructure (WebRTC, Daily.co, Deepgram)

What We Offer

✔ Early-stage equity in a healthcare AI company targeting a $12.6B combined TAM.

✔ Direct access to the CEO/Architect and CTO—no layers of management between you and the product.

✔ Real clinical impact—your code will power clinical decisions for patients and providers.

✔ Greenfield architecture ownership—you will shape foundational systems, not maintain legacy.