Evil Work
Evilworks offers PUFF, a data modeling platform that tracks data changes and provides full lineage for every number, making root cause analysis instantaneous and creating a complete audit trail. It addresses the inefficiency, security risks, and explainability gaps in traditional data workflows, supporting both human-led and AI-assisted processes. The platform is designed for enterprises where data errors can halt production or trigger regulatory fines.
- Artificial Intelligence
- AI Agents
- Data & Analytics
- Enterprise Software
- Software Only
Funding
Founders
Product
Problem
Working with data is slow, with business questions that should take two weeks often taking six months. When things go wrong, finding the cause can take days, and security risks from data leakage or malicious code injection are real. Unexplainable results carry regulatory risk with financial consequences up to 15,000,000 Euros.
Solution
PUFF is a data modeling platform that makes data analysis efficient, secure, and explainable. It tracks what changed in the data, saving compute and making root cause analysis often instantaneous. Every single number has full lineage information, creating a complete audit trail. PUFF removes the threat of malicious code injections, leaked data, and dangerous AI agent hallucinations, while confidential information is completely restricted. It enables both human-led and AI-assisted workflows without requiring an AI adoption, solving the same problems for humans and agents alike.
Target Audience
Primary customers are enterprises in financial services, manufacturing, and other regulated industries where data errors can cause production outages costing millions per day or result in regulatory fines.
Features
- Tracks data changes to save compute and enable instantaneous root cause analysis
- Provides full lineage information for every number, creating a complete audit trail
- Removes threats of malicious code injections, leaked data, and dangerous AI agent hallucinations
- Restricts confidential information completely
- Supports both human-led and AI-assisted workflows without requiring AI adoption
- Enables graph-based fraud detection by uncovering hidden network structures that row-by-row models miss