About Cupola

Built for people who need to know what is happening in their market.

Why we built this

Staying on top of a fast-moving market is a full-time job. Checking competitor sites, scanning newsletters, searching for funding announcements, reading analyst reports: it never ends, and most of it turns out to be noise.

Cupola started as a personal tool: a set of agents that watched a single market niche, filtered out the noise, and wrote a daily brief. It worked well enough that we built it into a product others could use. The same pipeline that powers our own research now runs for every workspace on the platform.

How the intelligence pipeline works

Every workspace runs a five-stage agent pipeline on a recurring schedule. Each stage hands off to the next.

Scout

Fetches content from your configured sources: RSS feeds, web searches for your keywords, competitor sites, and research repositories.

Curator

Scores each item for relevance to your niche and removes duplicates. Low-signal items are filtered before any AI analysis runs.

Analyst

Reads passing items and extracts structured records: named entities, themes, and key claims. These accumulate in your knowledge base.

Editor

Cross-checks claims for internal consistency, flags contradictions, and rates each source for credibility. The verification layer.

Synthesizer

Reads the full knowledge base and produces your daily brief: what changed, what it means, and what themes are building over time.

Team

Alexander Chacon

Founder

Data analyst and AI engineer at the University of Denver. Spent four years building intelligence systems for DU’s volleyball program before turning that same approach to market research and building Cupola.

hello@cupolahq.com