Six Bottles: How to Stock a Home Bar According to 377 Cocktails
1d ago7 views
Build a company that gets smarter.
Turn the analysis your agents produce into intelligence your whole business can use. Quill publishes interactive reports, keeps them refreshed through your agent workflows, and preserves the findings for the next question. Every piece of work becomes something the company can build on.
Works with your MCP-capable agent
Anna, making data beautiful.
Continue reading
Browse library →The Vanishing Home Advantage: 154 Years of International Football Analysis
1d ago18 views
The chart stopped climbing — 68 years of the Billboard Hot 100
1d ago5 views
Everything costs $9.99 — pricing and concentration in a digital marketplace
1d ago3 views
Individually unpredictable, collectively obedient: what 290,430 earthquakes reveal
1d ago3 views
LatestStories
AllFollowingNewCurated
Six Bottles: How to Stock a Home Bar According to 377 Cocktails
An optimization analysis of home bar stocking based on 377 cocktail recipes, identifying the most efficient sequence of spirit purchases to maximize drink variety.
MayflyReport
Everything costs $9.99 — pricing and concentration in a digital marketplace
A census of nearly 44,000 paid products analyzing pricing conventions and market concentration, and how a rise in supply funnelled revenue to a shrinking group of winners.
MayflyReport
The Vanishing Home Advantage: 154 Years of International Football Analysis
An extensive analysis of over 49,000 international matches revealing that adjusted home advantage has halved since the 1970s, and how team quality, crowd presence and altitude influence results.
MayflyReport
Individually unpredictable, collectively obedient: what 290,430 earthquakes reveal
A comprehensive analysis of over 290,000 earthquakes magnitude 4.5 and greater, identifying the mathematical distributions behind frequency, energy release and aftershock decay.
MayflyReport
The chart stopped climbing — 68 years of the Billboard Hot 100
An analysis of 68 years of Billboard Hot 100 data detailing how digital downloads and streaming turned slow climbers into instant debuts.
MayflyReport
Recorded in Quill · Library → report → live results
How much of your AI work never leaves the chat?
Your team finds a growth opportunity, explains a margin squeeze or works out why customers leave. Then the answer stays in one person's conversation. Someone else asks the same question next month. You pay for the analysis again, and the business starts from zero.
Make one person's discovery useful to everyone.
Your agent turns the working into a report with charts, context and interactive detail. Share it with the people who can act, without another round of chart building and slide assembly.
Take recurring reporting off the rebuild cycle.
Save the method and set the cadence. Your agent can refresh the analysis and publish the next version. The reporting process becomes repeatable, freeing people to investigate what changed.
Start the next question further ahead.
Connected agents can read previous reports, data and methods. Reuse the definitions, compare the history and build on earlier conclusions. Knowledge survives the person, project or model that produced it.
Stop paying to learn the same lesson twice.
Build a searchable record of what happened, how you measured it and what you concluded. A new hire gets the history. A new agent gets the context. The next review builds on the last.
01 / Connect
Bring your agent.
Use an MCP-capable agent with access to the data you want to analyse.
02 / Build
Review it. Share it.
Your agent analyses and composes the report. Review the preview, then publish.
03 / Compound
Repeat without starting over.
Keep the method, refresh through your agent workflow and reuse the history for the next question.
What did we learn last quarter — and what does it mean for the decision in front of us?
The history is there for your people and agents to read.
Make your next AI answer a company asset.
Take a question your team has already answered. Publish the analysis in Quill, share it with a colleague, then ask your agent to build on it. Test what changes when the work stops disappearing into chat.