Skip to content
Andrea Martinez Vernon

Experience

Highlights from each role. Happy to share a full CV — reach out onLinkedIn (opens in a new tab)

On this page

BenchSci

London, UK (remote)

  1. Senior Software Engineer, Data

    Dec 2024 – present

    Integrating heterogeneous biomedical data sources to power data-driven applications and data releases, in a globally distributed, interdisciplinary team.

    • Designed and shipped a Python ingestion platform that cut onboarding a new data source from 1–6 weeks to about one week.
    • Shipped a knowledge graph release that roughly tripled its association count, fully backward-compatible so no consumer team had to change a query.
  2. Deputy → Acting Engineering Manager, Knowledge Graph & Core Data

    May 2025 – Jul 2026

    Alongside the engineering role: covered engineering-manager duties for two teams at once — planning, 1:1s, performance and delivery — and managed a cross-team extraction of product data into a standalone pipeline through to its first production run.

    • Wrote the brief and OKRs for an ingestion-platform rebuild, negotiated the scope down from a full rewrite, and closed it on time.
    • Reintroduced team rituals and cross-team ceremonies, which other teams then adopted as a model.

BenevolentAI

London, UK

  1. Lead Bioinformatics Data Scientist / Squad Lead

    Jan 2024 – Dec 2024

    Led the team responsible for central data engineering, integrating 80+ internal and external datasets into data releases. Led the redesign of the central ETL framework so any engineer in the company could contribute ETL logic, with data governance built in.

    • Line-managed three data engineers, including supporting one into an ML and data science role that matched their interests.
    • Coordinated a data contract with the omics data-producing team, governing how their data was produced for the knowledge graph.
  2. Senior Bioinformatics Data Scientist

    Jan 2022 – Dec 2023

    Data engineering for the biomedical knowledge graph: developed the company's concept and relationship-type ontology and delivered graph updates. Squad Lead from Mar 2023.

    • Built Lagoon, a self-serve ingestion framework, and ran the workshops that let data engineers on other teams load data into the central data lake without our help.
    • Worked with a cross-functional team to improve drug-target safety assessment, delivering the data that could strengthen it through a new Elasticsearch index.
  3. Advanced Bioinformatics Data Scientist

    Nov 2019 – Dec 2021

    Data modelling and ETL systems for the knowledge graph behind the company's drug-discovery platform.

    • Designed ingestion and data models for sources including multi-omics data, Open Targets and ClinicalTrials.gov, with provenance and revisions built in.
    • Ran training workshops that let bioinformaticians and software engineers contribute directly to the knowledge graph.

Aigenpulse

Abingdon, UK

  1. Senior Computational Biologist

    Jun 2019 – Oct 2019

    Product owner for the CytoML Suite, a cytometry data management and analysis product.

    • Designed data workflows that assessed cytometry data quality and informed go/no-go decisions on downstream drug-discovery analysis.
  2. Computational Biologist

    Nov 2018 – May 2019

    Helped turn bespoke, customer-specific solutions into a scalable SaaS product taking life scientists from raw instrument data to decisions.

    • Talked to end users to surface cytometry flow-analysis requirements that had been missed, which were then built into the product.

University of Warwick

Coventry, UK

  1. PhD Researcher, Synthetic Biology

    Oct 2014 – Sep 2018

    Built software and lab tools to study how microbes share energy by exchanging electrons, as part of the EPSRC & BBSRC-funded Synthetic Biology Centre for Doctoral Training. Thesis: tool development for metabolic analyses in the context of thermodynamic constraints.

    • Developed MetQy, an R package published in Bioinformatics that turned KEGG's one-search-at-a-time web tools into bulk, automated analysis of the metabolic functions encoded in over 5,000 genomes.
    • Used MetQy to screen genomes for microbes likely to exchange electrons with electrodes, validated the search against known electroactive species, and narrowed roughly 3,500 candidates to 116 testable organism pairings.
    • Built a modular anaerobic electrochemistry platform (custom electrodes, 3D-printed parts, wiring and noise shielding) that held strict anaerobic conditions for over three weeks and ran a full 12-cell experiment at once, with four replicates per condition where most published studies use one.
    • Ran a 25-day experiment on the platform to test whether a sulphate-reducing bacterium and a methanogen could share energy through a wire: logged current continuously, characterised the electrodes, measured lactate and acetate by ion chromatography, and tracked control-culture growth by daily optical density.
    Lab bench with three sealed plastic containers of electrochemical cells stacked inside an open metal Faraday cage, wired to a potentiostat, a multiplexer and a monitor showing a current trace
    The anaerobic electrochemistry platform: sealed containers of electrochemical cells inside a Faraday cage, wired to a potentiostat and an 8-channel multiplexer.
    Sunburst chart with pathway classes in the inner rings and specific metabolic modules in the outer ring, each outer segment coloured by how complete that module is
    MetQy output: how complete each KEGG metabolic module is in one methanogen's genome, from broad pathway classes (centre) to specific modules (outer ring). Also published in Martinez-Vernon et al. (2018), Bioinformatics.

Projects

Human-in-the-Loop Agentic Résumé Pipeline

Sep 2026 – Oct 2026

Personal project, built with Claude Code as the coding agent.

  • Designed and built an open-source, human-in-the-loop agentic pipeline of Claude Code skills and subagents that mines job postings into an evidence-backed skills inventory, walks the candidate through each gap, drafts tailored résumés and critiques them automatically, accepting nothing the model proposes without human confirmation.
  • Created an interactive public demo that showcases the pipeline end to end on fictional personas: visitors approve, edit or reject each AI-proposed step, from CV to skills inventory to gap analysis to tailored résumé.

Try the demoRead the case studySee the repo on GitHub (opens in a new tab)

Skills and tools

Programming languages
Python, SQL, PySpark, R, Java, Bash
Data pipelines
Apache Airflow, Prefect, Apache Beam, Argo Workflows, Databricks, BigQuery
Knowledge graphs
Neo4j, SPARQL, RDF, OWL, LinkML, biomedical ontologies
Cloud and infrastructure
Google Cloud, AWS, Kubernetes, Docker, Terraform
AI
Retrieval-augmented generation, Claude API, agentic development workflows
Practices
Release management, privacy and compliance engineering, FAIR data
Leadership
Engineering management, technical leadership, stakeholder management, mentoring
Spoken languages
Spanish (native), English (proficient), German (intermediate)

Education

  • PhD, Synthetic Biology

    University of Warwick, UK · 2014 – 2019 (awarded Feb 2019)

  • MSc, Systems Biology (Distinction)

    University of Warwick, UK · 2013 – 2014

  • BSc (Hons), Biotechnology

    Universidad Autónoma de Querétaro, Mexico · 2008 – 2012

Selected publications