~/barcelona · senior director · ai enabling units

Shawn C. Kyzer

I build AI things that ship.

view experience say hello

shawn@bcn:~$ whoami
shawn-c-kyzer
shawn@bcn:~$ role --current
senior director · ai enabling units @ astrazeneca
shawn@bcn:~$ stack --building
["llm-agents", "rag", "doc-intel", "data-platforms"]
shawn@bcn:~$ ship --env regulated
✓ built. audited. in production.

15+: years across software, data, and AI
25+: engineers led across global teams
4: regions my engineering teams span

about

Demos are easy. Production is the job.

I lead AI enabling units at AstraZeneca in Barcelona — software, AI, and data engineering teams turning models into running systems: AI applications, data products, and the platforms underneath them, built side by side with research, IT, and business stakeholders.

The work I care about starts where the demo ends: architecture, model strategy, build-vs-buy calls, ML pipelines, infrastructure as code, budgets — and the evals, monitoring, and human-in-the-loop design that keep AI honest in a regulated domain.

llm agents & rag
document intelligence
data platforms & products
ml infrastructure
evals & observability
human-in-the-loop design
build-vs-buy strategy
gxp & regulated domains
team building

class ShawnKyzer:
    """The gap between a slick demo and a system
    people bet on — that's my lane."""

    def ship(self, idea: str) -> ProductionSystem:
        return pipeline(
            architecture, model_strategy, build_vs_buy,
            ml_pipelines, infra_as_code, budgets,
            evals, monitoring, human_in_the_loop,
        )  # governance is a feature, not a meeting

selected work

Proof over promise

# document intelligence

AI-assisted document authoring

Took an LLM-powered authoring assistant from architecture to production in a regulated environment — enterprise auth, monitoring, SLAs, and human-in-the-loop review included.

» architecture to production, end to end

# data platform

Early science data platform

Architected and shipped a GxP-compliant enterprise data hub for molecule, compound, assay, and biomarker data products — governance and lineage by design, not bolted on.

» sharp cut in manual data processing

# rag agent

Document intelligence agent

Led the first-place team at a Barcelona data engineering hackathon, building a retrieval-augmented agent that orchestrates document intelligence over a cloud data warehouse.

» 1st-place hackathon winner

resume

Experience

Senior Director, AI Enabling Units

AstraZeneca · July 2026—Present · Barcelona, Spain

Lead AI enabling units for AstraZeneca, spanning software, AI, and data engineering delivery.

Director, R&D IT — Data, Analytics & AI, Engineering

AstraZeneca · 2023—July 2026 · Barcelona, Spain

Directed and mentored distributed teams of 25+ Software Engineers, AI Engineers, and Data Engineers across the US, Europe, Guadalajara, and Chennai, delivering production-grade AI systems, data products, and enterprise platforms.
Led the global drug discovery data platform for the early science research domain, owning the delivery budget across FTE, outsourced service providers, and professional services.
Architected and shipped a GxP-compliant enterprise data hub of high-value data products — molecule, compound, assay, and biomarker — with governance, lineage, and provenance by design, sharply reducing manual data processing.
Championed an LLM-powered document authoring assistant, delivering end-to-end engineering from architecture through production deployment with enterprise authentication, monitoring, SLAs, and human-in-the-loop oversight.
Supported the first data engineering hackathon in Barcelona and led the 1st-place winning team, building a retrieval-augmented AI agent for document intelligence.
Established a new high-performing, diverse engineering team in Barcelona from the ground up, with clear career paths across software, ML, data, and platform engineering.
Served on the Barcelona IT Leadership Team as the primary R&D IT representative and Science & Innovation Ambassador, driving cross-functional partnerships and contributing to enterprise governance and architecture review boards.
Built the monitoring and documentation layer — model registry, performance tracking, decision audit trails — that makes AI approvable in a GxP environment.

Principal Data Engineer

Thoughtworks · 2020—2023

Served as Data & AI Platform Strategist for enterprise clients, developing multi-year platform strategies and technical roadmaps.
Led Data Mesh implementation from the ground up, including self-serve platform design, domain ownership, and federated governance.
Led a team building an SDK for data products so developers could discover, access, and integrate data products consistently.
Mentored Data & AI Engineers and delivered conference talks and training on data governance, data mesh, and ML platforms.

Senior Software Engineer

Skyscanner · 2019—2020

Led data modelling, reporting, and analysis across the platform stack as Data Champion, introducing Databricks reporting in CI/CD on AWS.
Led design reviews and production microservice implementations on AWS with Kubernetes orchestration.
Pioneered Kubernetes and Docker infrastructure migration while maintaining 90% code coverage across unit and integration tests.

Lead Solution Architect

Zurich Insurance · 2018—2019

Established an Architecture and Design Board for collaborative platform architectures and technical standards across business units.
Chaired architecture decisions under the Head of Enterprise Architecture for Data and Analytics.
Led Big Data and Data Science projects and evaluated AWS and Azure platform technologies for organisation-wide adoption.

Senior Principal Big Data Architect

ByteCubed · 2016—2018

Led cross-functional teams designing and deploying enterprise data platforms at the Pentagon under NIST and FedRAMP requirements.
Managed Terraform and Ansible deployments and data access controls across classified and unclassified environments through the ATO process.
Built ETL/ELT pipelines, data integrations, Docker containerisation, and cloud infrastructure provisioning with Python and Java.

2016       upside-travel   software dev / big data engineer
2015—2016  ikanow          lead dev / big data engineer
2014—2015  trusted.com     lead developer
2013—2014  booz-allen      software dev / data scientist
2012—2013  sapient         software developer
2010—2012  coventry/aetna  software dev, compliance mgmt

[mba]
school = "University of Maryland University College"
years  = "2012—2013"

[msit]
focus  = "software engineering"
school = "University of Maryland University College"
years  = "2009—2012"

[ba]
focus  = "philosophy"
school = "Swarthmore College"
years  = "2000—2004"

{
  "ai_ml": ["LLMs", "RAG", "agentic systems", "doc intelligence", "NLP", "MLOps/LLMOps", "PyTorch", "TensorFlow", "Hugging Face", "MLflow", "SageMaker"],
  "data": ["Snowflake", "Databricks", "Spark", "Hadoop", "ETL/ELT", "feature stores", "data mesh", "Power BI", "SQL"],
  "cloud_devops": ["AWS", "Azure", "GCP", "Terraform", "Ansible", "Docker", "Kubernetes", "CI/CD"],
  "languages": ["Python", "Java", "SQL", "Scala", "JavaScript", "C#"],
  "compliance": ["GxP", "GDPR", "NIST", "FedRAMP", "SAML/SSO", "IAM", "responsible AI"]
}

- [x] 2025 · AstraZeneca CEO Award
- [x] 2025 · AstraZeneca Exceptional Contributor Award
- [x] Keynote — GoTech World: Better Decisions Through Big Data & AI
- [x] Speaker — Open Data Science Conference
- [x] MLOps Coffee Sessions #116 — Scaling ML with Data Mesh
- [x] 2017 · ByteCubed Employee Recognition Award
- [x] 2016 · Ikanow Excellence Award

contact

Always up for talking shop

Based in Barcelona, Spain. I like comparing notes with people working on AI that has to survive the real world — regulators, auditors, production traffic. If that's your world too, the door's open.

shawn@bcn:~$ contact --say-hello

email → shawnkyzer@gmail.com linkedin → /in/shawn-kyzer-msit-mba-b5b8a4b github → /ShawnKyzer