LKQ Corporation

About the role

You’ll build the intelligence behind our automated content pipeline: expand and cluster search

topics, extract medical entities, infer user intent, triage risk (red/green), and generate structured

outputs that feed our auto-script templates. You’ll own accuracy, speed, and cost per script.

What you’ll do

● Design and ship the analyze(query) service: embeddings, clustering, entity/intent

classification, and risk flags.

● Build/maintain entity normalization to medical ontologies (RxNorm, SNOMED-CT/ICD-

10, MeSH).

● Implement risk triage rules (e.g., drug-drug interactions → clinical review queue) with

confidence scoring.

● Create and tune prompt chains / small models for auto-script scaffolding (hooks, VO

beats, CTA, citations).

● Stand up an evaluation harness (precision/recall/F1, calibration, cost/latency) and A/B

tests.

● Partner with SEO Lead Compliance to encode brand voice, disclaimers, and approval

logic.

● Collaborate with Data Eng on feature stores, vector DBs, data quality, and retraining

schedules.

● Monitor drift, errors, and red-flag rates; ship fixes fast (MLOps with versioned models).

What you’ve done

● 4+ years in applied NLP/ML (production systems, not just notebooks).

● Proficiency in Python, PyTorch/TF/JAX, Hugging Face, sentence-transformers,

spaCy. (or similar)

● Built text classifiers/NER and semantic search/embeddings in production.

● Experience with vector databases (pgvector, Pinecone, Weaviate) and FastAPI/Flask

services.

● Comfortable with ML Ops basics (MLflow/Weights Biases, CI/CD, containers).

● Strong grasp of evaluation (test sets, human-in-the-loop, active learning).

● Bonus: healthcare NLP, RxNorm/SNOMED/ICD-10, drug-interaction/contraindication

logic, LLM prompt engineering, distillation/LoRA.

How we’ll measure success (first 90 days)

● v1 analyze() API live with ≥0.80 F1 on high-support intents; entity normalization ≥0.95

accuracy.

● Red-flag recall ≥0.95 (we’d rather over-catch early).

● Cost/latency targets met for daily batch runs; clear, self-serve evaluation dashboard.

Tech you’ll touch

Python, PyTorch, Hugging Face, sentence-transformers, spaCy, FastAPI, Postgres/pgvector or

Pinecone, Airflow/Prefect, MLflow/WB, Docker.

Compensation
$140,000 - $190,000 Per Year

Job Type: Full-time

Pay: $140,000.00 - $190,000.00 per year

Benefits:

• Dental insurance
• Health insurance
• Paid time off
• Vision insurance

Application Question(s):

• Will you now or in the future require sponsorship for employment visa status (e.g., H-1B) to work in the United States?

Work Location: In person