Dialectica Origin: a data platform built on Webflow Cloud

The challenge

Dialectica runs Origin, a deal intelligence platform used by private equity, venture capital, and corporate strategy professionals. They wanted to bring a slice of that product to the public web: a free, searchable version of their company database, living as a subfolder on their main domain.

The goal was twofold. Test whether they could sell data through a freemium model. And boost organic search by getting hundreds of company and industry pages indexed by Google.

The problem: they had over 10.000 company records to show. Webflow’s CMS caps at 10.000 items, and even if it didn’t, the data needed complex relational fields, advanced filtering, and real-time updates that standard Webflow CMS couldn’t handle. This wasn’t a blog or a portfolio. It was a database-driven product.

The solution

We were already handling Webflow maintenance for Dialectica on a monthly retainer when they asked if it was possible to bring Origin data into their site. The CMS route was out immediately. We needed a database.

Webflow Cloud had just come out of beta, and it was the right fit. It lets you deploy a full-stack application as a subfolder of an existing Webflow site, powered by Cloudflare Workers under the hood. Dialectica keeps their marketing site in Webflow. Origin runs as an Astro app inside it. Same domain, same navigation, completely different architecture underneath.

Astro handles the frontend with server-side rendering on the edge. Company pages, industry pages, and programmatic SEO pages are all rendered on request. React components handle the interactive bits: multi-select filters, range sliders, sorting, pagination, and a lead capture form.

Neon provides the serverless Postgres database. Dialectica exports company data as JSON files and we import them into Neon using Python scripts. The serverless driver works natively with Cloudflare Workers, so queries run at the edge with no cold starts.

Technical approach

Data masking for freemium. Dialectica wanted to tease their data without giving it away. We built two layers of protection: server-side masking transforms real values before they reach the browser (revenue figures become $5••M, investor names get obscured), and a CSS blur filter adds visual obfuscation on top. Scrapers only see masked data. Humans see enough to want more. A lead capture form triggers based on scroll depth and time on page.

Programmatic SEO at scale. Every company gets its own page. Every industry gets a landing page. Then we generate combination pages: industry + location, industry + ownership type. Each with proper meta tags, JSON-LD structured data, and a dynamic XML sitemap. The result is hundreds of indexable URLs, all generated from the database.

AI-assisted development. The entire build was done by one developer, part-time, in under two months. That timeline was only possible because AI tools handled much of the implementation work. This was a first Astro project, a first Webflow Cloud project, and a first serverless Postgres project. Claude Code helped write TypeScript logic, Python data migration scripts, and debug edge runtime compatibility issues throughout.

The results

The MVP is live at dialectica.io/origin. Google has indexed the company and industry pages, and some are already ranking on the first page for relevant queries.

The architecture works. Adding new companies means importing a JSON file. Adding new page types means creating a new Astro route. The current 931 companies are a starting sample; the system handles 10.000+ with no structural changes needed.

Dialectica got a public-facing data product without rebuilding their main site. And the project proved that Webflow Cloud can power real data applications, not just marketing pages.

For the full technical deep-dive into the architecture, read Scaling beyond Webflow CMS: a hybrid architecture.