# Elena Marquez

Headline: Senior AI Engineer | LLM Systems & Machine Learning

- Email: elena.marquez.ai@gmail.com
- Phone: (415) 555-0148
- URL: https://elenamarquez.dev

Location: 500 Terry A Francois Blvd, San Francisco, California, United States, 94158

Profiles: 
- GitHub: [@elena-marquez](https://github.com/elena-marquez)
- LinkedIn: [@elena-marquez-ai](https://www.linkedin.com/in/elena-marquez-ai)
- Medium: [@elena-marquez-ai](https://medium.com/@elena-marquez-ai)

## Basics

AI Engineer with 6+ years of experience designing, evaluating, and deploying production machine learning and large language model systems. I turn ambiguous product goals into measurable models, reliable inference services, and evaluation frameworks that teams can trust.

- Built retrieval-augmented generation and agentic workflows for healthcare and enterprise search, improving answer quality while reducing latency and cost.
- Led model evaluation, prompt optimization, and guardrail design across multiple product teams, with a focus on safety, observability, and reproducibility.
- Partnered with product, data, and platform engineers to ship ML features from prototype to production with clear ownership and monitoring.


## Education

### University of Washington

Master, Computer Science and Artificial Intelligence, Score: 3.9, Sep 2017–Jun 2019

URL: https://www.washington.edu/

Courses: Machine Learning, Deep Learning, Natural Language Processing, Probabilistic Graphical Models, Distributed Systems, Statistical Inference, Human-Computer Interaction

Summary: 
- Completed graduate coursework and research at the intersection of machine learning, NLP, and human-centered AI.
- Built reproducible experiments for text classification and retrieval, including error analysis and ablation studies.
- Presented research findings to academic and industry audiences and collaborated across disciplines.


## Work

### Senior AI Engineer

Aether Labs, Mar 2023–Present

URL: https://www.aetherlabs.ai

Keywords: LLM Systems, RAG, Evaluation, Inference, Mentorship

Summary: 
- Lead the design and deployment of LLM-powered search and assistant features used by enterprise customers.
- Built a retrieval-augmented generation pipeline with hybrid search, reranking, and citation-aware responses, reducing unsupported claims by 38%.
- Created an offline and online evaluation stack covering relevance, groundedness, latency, and cost, enabling weekly model and prompt releases.
- Partnered with platform engineers to productionize inference services with batching, caching, and observability on Kubernetes.
- Mentored two AI engineers and established review practices for prompt changes, data contracts, and model risk assessments.


### Machine Learning Engineer

BrightPulse Health, Jul 2019–Feb 2023

URL: https://www.brightpulsehealth.com

Keywords: Healthcare, NLP, Risk Models, AWS, Monitoring

Summary: 
- Developed clinical risk and triage models using structured EHR data, improving recall for high-risk patients by 17% in prospective validation.
- Built NLP pipelines to extract signals from unstructured notes, with de-identification, labeling, and active learning workflows.
- Deployed batch and real-time inference services on AWS, with data quality checks, model monitoring, and rollback procedures.
- Worked with clinicians and compliance teams to document model limitations, intended use, and bias mitigation steps.
- Automated feature pipelines in Airflow and Spark, cutting weekly model refresh time from two days to four hours.


### AI Research Intern

VectorForge AI, Jun 2018–Sep 2018

URL: https://www.vectorforge.ai

Keywords: Embeddings, Semantic Search, Research

Summary: 
- Prototyped sentence embeddings and retrieval baselines for semantic search over technical documentation.
- Ran ablation studies on negative sampling and reranking, documenting findings in an internal research report.
- Contributed reusable evaluation scripts and data preparation notebooks to the research team.


## Languages

- English: Native or Bilingual Proficiency, Keywords: Technical writing, Conference talks
- Spanish: Full Professional Proficiency, Keywords: Bilingual collaboration
- French: Elementary Proficiency, Keywords: DELF A2

## Skills

- Machine Learning: Expert, Keywords: PyTorch, TensorFlow, scikit-learn, XGBoost, Statistical Modeling
- LLM & Generative AI: Advanced, Keywords: Transformers, LangChain, RAG, Prompt Engineering, Fine-tuning, Agents
- Data Engineering: Advanced, Keywords: Python, SQL, Apache Airflow, Apache Spark, dbt, Snowflake
- Cloud & MLOps: Advanced, Keywords: AWS, GCP, Docker, Kubernetes, MLflow, Terraform
- Software Engineering: Advanced, Keywords: FastAPI, TypeScript, REST APIs, CI/CD, Testing, System Design

## Awards

### AI Platform Innovation Award

BrightPulse Health, Nov 2022

Summary: 
Recognized for leading the design of a clinical NLP platform that reduced manual chart review time by 30% while maintaining auditability and clinician trust.

### Graduate Research Excellence Award

University of Washington, May 2019

Summary: 
Awarded for research on retrieval and evaluation methods for domain-specific language models, including a published workshop paper and reproducible benchmark.


## Certificates

### AWS Certified Machine Learning - Specialty

Amazon Web Services, Aug 2022

URL: https://aws.amazon.com/certification/certified-machine-learning-specialty/
### Google Cloud Professional Machine Learning Engineer

Google Cloud, Jan 2023

URL: https://cloud.google.com/certification/machine-learning-engineer

## Publications

### Evaluation-Driven Development for Retrieval-Augmented Generation

Proceedings of the Conference on Applied Machine Learning, Oct 2023

URL: https://dl.acm.org/doi/10.1145/3612345.3612399

Summary: 
- Proposes a practical evaluation loop for RAG systems that combines retrieval diagnostics, groundedness checks, and human review.
- Shares lessons from deploying the framework across customer support and healthcare search use cases.


### Calibrating Clinical Risk Models Under Distribution Shift

Journal of Healthcare AI, Mar 2021

URL: https://doi.org/10.1016/j.jhai.2021.03.004

Summary: 
- Studies calibration drift in clinical risk prediction models when patient populations and care pathways change.
- Introduces monitoring signals and retraining triggers that helped maintain model reliability in production.


## Projects

### RAG Evaluation Harness

Open-source toolkit for reproducible evaluation of retrieval-augmented generation pipelines, Jan 2024–Present

URL: https://github.com/elena-marquez/rag-eval-harness

Keywords: RAG, Evaluation, LLM

Summary: 
- Provides pluggable retrievers, rerankers, and judges for measuring relevance, groundedness, and answer completeness.
- Includes synthetic query generation, regression tests, and dashboards for comparing prompt and model changes.


### Clinical Note Summarizer

Privacy-preserving summarization service for clinician-authored notes, Feb 2022–Nov 2022

URL: https://github.com/elena-marquez/clinical-note-summarizer

Keywords: Healthcare, NLP, Summarization

Summary: 
- Built a de-identification and summarization pipeline with human-in-the-loop review and audit logging.
- Evaluated summaries for factual consistency, coverage, and clinician acceptance before a limited pilot.


## Interests

- Open-Source AI: Evaluation, Reproducibility, Agents
- Outdoors: Trail Running, Climbing, Cycling
- Music: Jazz Piano, Music Production

## Volunteer

### Women in Machine Learning

Mentor, Jan 2021–Present

URL: https://wiml.org

Summary: 
- Mentor early-career engineers and researchers on ML project scoping, interview preparation, and career growth.
- Review workshop submissions and support community events focused on inclusive AI research and practice.
