Spacecubed: Turning Hours of Manual Company Research Into One Pasted URL and a Few Seconds

Brief

Spacecubed’s Echo platform helps foundations and boards manage grants, applications and impact reporting, and every company added to the platform needs a complete profile: contact email, phone number, physical address, LinkedIn, map location, logo, sector. Building those profiles meant staff manually browsing company websites, trawling Google and copying details across by hand. We built an AI extraction agent that does the whole job from a single URL. A user pastes a company’s website address, and the agent reads the site, explores its other pages, fills the gaps with targeted Google searches, locates the matching Google Maps object, and returns a complete, unified company profile in seconds. It was delivered as a private npm package that slotted straight into Echo’s existing codebase.

I have used Sam and Sean’s services multiple times for AI related software integrations. I have found them to be extremely capable, great communicators and deliver exceptional results. I would highly recommend them!

Nate Sturcke, Spacecubed

Problem

Every platform that holds company data faces the same grind: someone has to put that data in. For Spacecubed’s Echo platform, that someone was their own staff, manually assembling company profiles one field at a time. Find the website, hunt for the contact email, dig out the phone number, search Google for the address, track down the LinkedIn page, match the location on a map, then repeat for the next company. A single complete profile could mean navigating multiple websites and search results, and at scale the hours stacked up fast. Worse, manual data entry is exactly where errors live, a transposed digit in a phone number, an outdated address, the wrong company’s LinkedIn. Spacecubed needed the research done by software: something that could take nothing but a URL and hand back a complete, accurate company profile, ready to drop straight into Echo.

Solution

Figure 1: Diagram of the AI company data extraction agent.

A user pastes a company website URL into Echo, and our AI agent takes it from there. It reads the site, intelligently explores the pages where company details actually live, such as /about, /contactus and /ourteam, and compiles everything into a structured profile. Where the website comes up short, the agent searches Google for the missing pieces, cross-checking against data it has already extracted to make sure it’s matching the right company, then returns the lot in seconds.

• One input, complete output: a single URL returns name, description, contact email, phone, address, LinkedIn, socials, logo, sector and a ready-to-use map location
• Explores like a human, works like software: the agent doesn’t stop at the homepage, it visits the about, contact and team pages where the real details hide
• Smart about identity: when searching Google for missing information, the agent cross-checks against data it has already extracted, so it never confuses two similarly named companies
• Maps integration built in: locations come back as Google Maps place objects, matching the integration Echo already had, so they plug straight into the platform
• Honest about gaps: anything genuinely unavailable returns as null rather than a guess, protecting the platform’s data quality
• Delivered to fit their stack: shipped as a private npm package that Echo’s team integrated directly, no new infrastructure, no platform rebuild

Result

Echo users now receive a unified company profile, name, contact email, phone number, address, LinkedIn and map location, within seconds of pasting a URL. The hours of manual research per profile are gone, and with them the copy-paste errors that manual entry invites. For Spacecubed, the win compounds with every company added: profile creation is no longer a bottleneck on growing the platform, and staff time that went into browser tabs and Google searches now goes into the platform itself. The build also gave us deep familiarity with Echo’s codebase, paving the way for future AI features on the platform.

Under the Hood