All Projects
Automation1 March 2024

Financial Data Extraction Pipeline

An automated bot that scrapes the FCA National Storage Mechanism, identifies company AGM and GM filing dates, and structures them into a live database updated daily.

PythonOpenAIPydanticpandasAWS Lambda

Outcome

Monitors hundreds of listed companies continuously. Zero manual checking of the NSM required.

Live Demo

850+

Companies Monitored

247

Filings Processed Today

0

New AGM/GM Dates Found

filings.fca.org.uk/nsm/nationalstoragemechanism
Filing IDCompanyTypeDate
UK6001234BarclaysAGM01 Apr 2025
UK6001891AvivaGM03 Apr 2025
UK6002104LegalAGM07 Apr 2025
UK6002398SchrodersAGM10 Apr 2025
UK6002511M&GGM12 Apr 2025
UK6002687HargreavesAGM14 Apr 2025
UK6002790abrdnAGM16 Apr 2025
Extracted Database
CompanyEventDateRes.
awaiting extraction...

The Challenge

Monitoring AGM and GM dates for hundreds of listed companies required manually checking the FCA National Storage Mechanism for new filings. This was done sporadically and inconsistently, meaning important meeting dates were sometimes missed or flagged late, impacting voting preparation timelines.

How It Was Built

Built an automated scraper that runs daily against the FCA NSM, identifying new AGM Notice and GM Notice filings by company name and filing type. When a new filing is detected, an LLM-based extractor parses the document to identify the meeting date, venue, number of resolutions, and record date. Extracted data is written to a structured database with deduplication and change detection. AWS Lambda handles scheduling and execution.

The Result

The system now monitors 850+ listed companies continuously with zero manual checking required. All AGM and GM dates are captured within 24 hours of filing and structured data is available for downstream use in voting calendars and engagement workflows.

Technology Stack

PythonOpenAIPydanticpandasAWS Lambda