How We Work
Our Technology
The tools, data sources, and systems that power our intelligence operation.
MARKET & FINANCIAL DATA
- —Yahoo Finance
- —OpenBB — open-source Bloomberg alternative for market data, financial statements, and equity research
- —Company Investor Relations Pages
- —Annual Reports & Sustainability Disclosures
REGULATORY & PUBLIC RECORDS
- —CSRD Filings & SFDR Classifications
- —EU Taxonomy Alignment Data
- —SEC EDGAR
- —Eurostat · ECB · National Regulatory Databases
- —Enforcement & Litigation Records
RESEARCH, ANALYSIS & MONITORING
- —OpenBB — quantitative data analysis and financial modelling
- —Custom AI Agents — proprietary agents built for specific research tasks across our three analytical lenses
- —Large Language Models — multiple LLM providers used across research, drafting, and synthesis workflows
- —Perplexity AI — real-time source monitoring and research synthesis
- —Google Alerts — keyword tracking across markets, regulation, and ESG developments
- —Academic & Industry Publications
DESIGN & VISUAL PRODUCTION
- —Microsoft 365 — report writing, editing, data analysis, data visualisation
- —PowerBI — data analysis & data visualisation
- —Adobe Creative Suite — report design, data visualisation, brand assets
- —Figma — layout design, web and report templates
- —Canva — social content, presentation decks, quick assets
Editorial Position
Our AI Usage Policy
We use AI extensively across our workflow. Here is where it operates — and where it does not.
AI assists our team with source monitoring, data aggregation, draft structuring, and synthesis across large volumes of information. We also deploy custom-built AI agents for specific, repeatable research tasks within our intelligence pipeline.
What AI does not do at Vigilis Impact Lab is draw conclusions, make editorial decisions, or determine our analytical position on any topic. Every report, score, and published finding is reviewed, stress-tested against the VIGIL Framework, and cleared by a human analyst before it reaches our audience.
The 20-Minute Rule applies to all AI-assisted content — no output enters publication without a minimum of 20 minutes of human review, fact-checking, and editorial judgment applied on top of it.
We use AI as leverage — not as a replacement for the thinking that makes our work credible.