Most companies pitching for government contracts face the same maddening cycle: hundreds of hours crafting proposals, then watching that hard-won expertise evaporate the moment a deal closes. Henry Brogan and Tyler McCarthy, childhood friends from Manchester who built their first startup while still at university, think they've found a way to break it.
Their company, BidScript, closed an $800,000 funding round—a follow-on that pushed its total pre-seed haul past the $1 million mark. The financing came from NPIF II – PXN Equity Finance, managed by PXN Ventures under the British Business Bank's Northern Powerhouse Investment Fund II umbrella, with fresh backing from SFC Capital joining the mix.
Editor's note: The draft referenced a funding date of July 2, 2026, which appears to be a placeholder or error. We've generalized the timeline pending confirmation of the actual close date.
What BidScript is selling isn't flashy—there's no consumer app, no viral growth loop. It's infrastructure for the decidedly unsexy world of tendering, where companies compete for lucrative public and private sector contracts by submitting exhaustive written proposals. But Brogan and McCarthy reckon the opportunity is bigger than it looks. Early adopters in the UK, US, and Middle East have reported win-rate jumps of up to 50%, according to the company, though independent verification of these results is not yet available.
Institutional Memory, Automated
The pitch hinges on what the founders call a structural inefficiency. Every time a firm bids on a new contract, teams dig through filing cabinets (literal or digital) hunting for relevant case studies, technical specifications, and past proposal language. More often than not, they just start over.
BidScript's answer is something it dubs the "Knowledge Engine"—a system that hoovers up entire documents, from old bids to internal reports, then uses semantic chunking and vector embeddings to retrieve the right fragments when a new opportunity surfaces. The platform plugs into SharePoint, OneDrive, Google Drive, HubSpot, Salesforce, and workplace chat tools like Teams and Slack.
An AI agent—marketed alternately as "BidBot" and "Biddy," depending on which page of the company's site you land on—handles the grunt work: research, retrieval, drafting, complete with citations. It's retrieval-augmented generation (RAG) applied to a compliance-heavy, high-stakes domain where generic chatbots tend to stumble.
"General-purpose AI does not meet the standards required for high-stakes public and private sector tendering," Brogan said in a statement. "It's a niche that demands deep expertise."
Whether BidScript has cracked that niche remains an open question. The company name-drops customers like CHAP, PeopleScout, DDS Group, and Enviro Technology Services on its pricing page, but detailed testimonials are scarce.
From PraeSeed to Follow-On
BidScript incorporated in April 2024. By late that year, it had secured £200,000 through PXN Ventures' PraeSeed program, a cohort-based initiative backing early-stage startups across Northern England. The fund has a mandate to deploy £6 million across 35 companies by 2028.
Brogan studied economics with finance at Loughborough University; McCarthy concentrated on AI and data science at Manchester Metropolitan. The company operates out of Altrincham, a suburb in Cheshire, and both founders hold between 25% and 50% voting rights, according to filings with Companies House.
This latest round will fund product development, hiring, and geographic expansion beyond the UK. BidScript's marketing materials tout a discovery engine covering more than 740 sources across 120 countries, surfacing some 10,000 new contract opportunities daily. Target sectors include construction, engineering, IT, education, and healthcare—industries where public procurement budgets run deep.
A Crowded Field, Perhaps More Than Expected
The tender-management software space isn't empty. Loopio and Responsive (formerly RFPIO) have been selling response-management platforms for years, and both have layered AI features into their products as large language models matured. What BidScript is betting on is that purpose-built tooling—AI designed from the ground up for this workflow—will outperform retrofitted features bolted onto legacy systems.
Harry Manley of PXN Ventures and Ed Stevenson of SFC Capital both emphasized the appeal of applying narrowly tailored AI to high-compliance environments like public procurement, where mistakes carry reputational and financial risk.
Still, the company will need to move quickly. The AI tooling landscape shifts by the quarter, not the year, and competitors with deeper pockets are paying attention. BidScript's next financial accounts, when filed, will offer the first public snapshot of how revenue is tracking against the founders' ambitions.
For now, Brogan and McCarthy are betting that the unglamorous work of automating government bids is exactly the kind of problem where narrow AI beats general-purpose models. And where a couple of friends from Manchester, armed with a million dollars and a theory about institutional memory, might just carve out a defensible niche.
