Engagements often start with an audit. We assess capabilities and map common workflows. We then develop recommendations and identify quick wins alongside longer-term opportunities and risks.
AI Systems for
Health Economics &
Outcomes Research
We advance AI capabilities in HEOR by developing new tools, workflows and evaluation methods.
Our thesis
HEOR is about to undergo a fundamental transformation.
Organising AI systems for extreme reliability and performance used to be prohibitively difficult. Advances in the implementations and in the models are eliminating those barriers.
We welcome this change and want to shape it responsibly, with human agency at its centre.
Offerings
We work with HEOR teams on AI adoption, custom tools and end-to-end research workflows.
We advise teams at different stages in their AI journey, from initial exploration through to full restructuring of workflows and adoption of AI-native systems.
We develop AI tools for tasks ranging from data extraction and visualisation to multi-step workflows and model QC. Tools include rigorous evaluations, provenance and full observability, and can be deployed within your infrastructure.
Our core offering is developing rigorous end-to-end workflows for complex tasks, such as SLRs, NMAs, CE-model review or model country-adaptations.
Workflows are built on evals and test sets to validate performance, and are designed to augment human judgment, not replace it.
Client stories
AI Integration
A UK HEOR consultancy sought help getting started using AI and agentic systems.
We conducted interviews and mapped core workflows. We then ran workshops and developed a plan for reorganising processes around AI capabilities.
AI use increased threefold and the team reported clear productivity gains, particularly in modelling.
Advisory & Tooling
A German consultancy, specialised in RWE, wanted support with scientific writing and building AI workflows.
We ran a half-day introductory session and co-developed initial prototypes for reviewing and drafting study reports, supported by weekly check-ins.
The team is now reshaping core workflows and actively developing their own AI solutions.
Task Automation
An NGO that delivers thousands of direct cash transfers spent 2-3 days each month on mandatory reporting tasks.
Over the course of three one-week sprints, we identified bottlenecks and implemented a pragmatic AI solution within existing tools.
Many reporting tasks are now effectively automated, freeing up staff time for higher-value work.
AI-native infrastructure
Current systems have been designed for humans. AI-native workflows require rethinking information infrastructure.
A knowledge graph of EuroQol-funded research: We built an automated pipeline to identify, retrieve and ingest data from 1,024 EuroQol projects and 797 publications into a knowledge graph. It connects researchers, methods and outcomes in one queryable resource, enabling entirely new ways of AI-driven meta-research. View project
Further resources
Covering 826 appraisals
Linking IQWiG and GBA cases
Automated study replication
MCP server for AI peer review
Evaluation-driven workflows
We believe evaluations will be central to effective and rigorous AI-native research workflows.
Further resources
Team

Paul Schneider MD, PhD
Founder and CEO
Paul combines expertise in medicine, health economics and research software engineering. He worked on AI for Science at the Foresight Institute's Berlin AI Node and previously co-founded Dark Peak Analytics Ltd., where he led projects with top-10 pharma companies, NGOs, and the WHO. In his academic work, he developed novel health valuation methods, created tools routinely used in NICE technology appraisals, and published 50+ peer-reviewed papers.
We bring together a team of exceptional researchers, engineers and product managers with deep expertise in health economics, market access, statistical methods and advanced AI systems.





Contact
We welcome enquiries about AI in HEOR, project opportunities and research collaboration.
or email us at contact@abundanceds.com