Tracking Funding, Jobs and News Signals Automatically

Why signals matter

Companies reveal their plans through public events. A funding round suggests growth and budget. A burst of job listings suggests expansion. News coverage can signal a launch, a merger or a restructuring. Sales and research teams value these signals, but collecting them by hand across thousands of companies is not realistic. We built a tool that does it automatically.

What it does

The tool uses search engine results for web, jobs and news queries to discover UK startup funding rounds, job listings and company news. It then resolves each discovered company to its official Companies House registration number and stores everything in a database. It supports broad discovery across the whole UK and targeted searches for a specific list of companies.

The hard part: knowing which company you found

Search results mention company names, not registration numbers, and names are ambiguous. Two companies can have similar names, trading names differ from registered names and press articles use shortened forms. Resolving a name to the right official record is the step that makes the data trustworthy and joinable with other systems. We match carefully, deduplicate by company and keep the resolved registration number on the record so that funding, jobs and news all attach to the same entity.

Broad discovery versus targeted runs

Broad discovery finds new things the user did not know to look for. Targeted runs check a particular list, such as a sales team's prospects. We tag rows so the two can be separated, and each targeted run keeps its own identifier, which makes results easy to review and export.

Engineering details that matter

• A layered structure: web interface, batch processors, pipeline logic and scrapers, so each part can be tested and changed on its own.
• Schema creation protected by a database lock so several server processes starting together do not collide.
• Lightweight migrations for adding columns as the tool evolves.
• A simple web interface with login, run controls and exports.
• Support for a local file database for development and Postgres in production.

Respecting sources and limits

Automated discovery should respect the terms and rate limits of the services it uses. We use a search API rather than scraping pages aggressively, schedule runs sensibly and store source links so every fact can be traced back to where it came from.

Ideas for your business

Monitoring competitors, suppliers, tenders, regulations or market news can all use this pattern: discover, resolve to a canonical entity, store and alert. We can build a monitoring tool tailored to your sector.