Frekil

Frekil is a platform that automates real world evidence generation for life sciences organizations. A researcher describes a clinical question in plain language and the platform's AI agents carry it through an eight stage epidemiological pipeline: literature review, question clarification, cohort building, causal diagram construction, statistical analysis plan generation, data extraction, plan execution and report writing. The result is a complete study run on the organization's own clinical data, with a person reviewing and correcting the agents at each stage. It is designed for pharmaceutical and biotech companies, contract research organizations, epidemiologists, HEOR and market access teams, post market safety groups and payers. The platform connects to electronic health records, claims, registries and proprietary datasets through Databricks, Snowflake and the major cloud platforms. It generates statistical code in R, Python and SAS, and its architecture keeps patient level data separate from the AI models so clinical data is never exposed to them. The vendor describes the agents as self improving, learning from each completed analysis, and states that the infrastructure is HIPAA and GDPR compliant. Pricing is not published. Frekil offers per study and platform pricing models, and prospective customers contact the company for a quote.