Artificial Intelligence Engineer

AI Engineer

1) Organization Overview (Concise & Neutral)

A fast‑growing oncology‑focused organization is reinventing how clinical trials operate by integrating them tightly with real‑world clinical practice. Cross‑disciplinary teams across healthcare, engineering, AI, and regulatory domains work in a Human‑in‑the‑Loop (HITL) model to deliver regulatory‑grade outcomes that expand trial access and accelerate cancer drug development.

What’s Different

* Clinical trials are embedded within clinical practice—not run in parallel.

* Hybrid model blending expert abstraction, AI/NLP/LLMs, and EMR integrations.

* Strong commitment to rigorous testing, validation, privacy, and ethical AI.

* Mission‑driven culture with high collaboration and urgency.

Why Join Now

* AI is essential to scaling the business—impact is immediate and visible.

* Opportunity to build an end‑to‑end applied AI stack powering clinical teams.

* Work across LLMs, CV, and multimodal use cases with strong platform engineering partners.

2) Role Overview — AI Engineer

Purpose

Design and deliver applied AI systems (LLMs/CV/multimodal) that automate clinical variable extraction and clinical note generation. Work includes repeatable validation, robust documentation, and HITL feedback loops. Strong emphasis on data engineering and backend rigor to ensure model usability and efficiency.

Focus Areas

* LLM development and LLM‑Ops for text extraction and structuring; some CV/multimodal.

* Rapid prototyping with high software‑engineering hygiene.

* Statistical validation plans, experiment design, and metrics ownership.

* HITL workflow development to reduce manual QA and improve throughput.

* Collaboration with Platform Engineering for production alignment.

* Comprehensive documentation: datasets, experiments, model cards, QA audits.

* HIPAA‑aligned safeguards and compliant AI practices.

Core Responsibilities

* Build AI models and pipelines across EMR/EHR, imaging, and clinical documents.

* Translate ambiguous clinical requirements into measurable ML objectives.

* Define metrics, design experiments, and estimate/model error.

* Lead interim QA audit processes and evolve toward AI‑assisted QA.

* Partner with data/platform engineers on scalability, data flow, and observability.

* Champion code quality, experiment tracking, reproducibility, and knowledge capture.

Expected Impact (6–12 Months)

* Deliver validated AI components for abstraction and note generation.

* Meaningfully reduce manual QA workload through HITL optimization.

* Standardize testing and documentation frameworks.

* Establish efficient PySpark/SQL/Postgres data‑manipulation patterns that accelerate iteration.

3) Product & AI Context

In‑Flight Work

* LLM‑powered abstraction and clinical note generation with HITL.

* Auditing and validation pipelines.

Mandate

* Drive productivity gains for internal labeling/abstraction teams.

* Build the full applied‑AI stack: model development, context engineering, QA automation, interfaces.

Productionization

* Deployment owned by Platform Engineering, but AI Engineers write scalable, integration‑ready code.

Data Gravity

* Data engineering is a major part of the role: PySpark, SQL/Postgres, query optimization, cloud data tooling.

4) Ideal Candidate Profile

Background

* MSc/PhD in CS, EE, Applied Math, Stats, Physics, or equivalent depth via experience.

* 2–5+ years in AI/ML engineering or applied data science.

* Healthcare or clinical workflows experience strongly preferred; oncology a plus.

Technical Must‑Haves

* Expert‑level Python + strong software engineering practices.

* Deep learning experience with PyTorch or TensorFlow (LLMs and/or CV).

* Data engineering: PySpark, SQL, Postgres, data modeling, query tuning.

* Cloud data platforms (Databricks, S3/Snowflake/Azure/GCP).

* Experiment design, statistical validation, and error analysis.

* HITL lifecycle design and feedback integration.

Nice‑to‑Have

* Additional languages: R, Java, C++.

* MLOps fundamentals (versioning, lineage, CI/CD).

* Prior oncology or clinical trials exposure.