Sloppod › about
About Sloppod
A thread with four hundred comments has maybe five arguments in it. Sloppod finds them, gives each one a voice, and lets them argue at the weights the thread actually held.
Why it exists
Reading a long thread is work, and most of it is repetition. Summarising one is worse: a summary flattens a disagreement into a paragraph that nobody in it would sign. What is actually interesting about a busy thread is the shape of the disagreement - how many positions there really are, and how much of the room stood behind each one.
So this does not summarise. It finds the clusters, gives each a speaker, and weights their airtime by their real support. A thread that went ninety-ten produces an episode that sounds ninety-ten. The parts it will not do - describe a photograph, fill a time budget, report a vote a platform does not have, name anybody - are listed on the front page, because they are the reason to trust the rest.
Who made it
Sloppod is built by Refleksion · Planlægning · Handling, a small Danish company (CVR-no. DK45818659) that builds quiet, opinionated tools. The three words in the name are Reflection, Planning and Action - the loop the other products are shaped around.
Independent and self-funded. No investors, no growth hacks, no dark patterns. The way this makes money is people paying for software they use.
Every episode costs real money to make - vision, four language-model calls and a voice clip per turn - which is why sign-up is closed while it is small, and why there is a waiting list rather than an open door.
Other sites we have built
You might like these too. All from Refleksion · Planlægning · Handling.
- Taskpile A task list your AI assistant can actually reach.
- Foodpile Track what you eat just by telling your AI chat.
- Jotpile early beta Catch every thought fast, find it again later.
- GeoReady Find out what AI really says about your brand.
- Membrr Run your club without the paperwork.
- Vidde Slip away into a quiet corner of space.
Getting hold of us
Write to inbox@sloppod.app. A person reads it.
What this holds about you, if you have an account: the address you signed in with, the episodes you made and the threads they came from. Nothing else. There is no tracking on this site — no analytics script, no third party, no cookie but the one that signs you in, and nothing that follows you from one page to the next. You can check that by reading the page source, and there is not much of it to read.
What does exist is what exists for every website: the server's own request log. Cloudflare keeps it for seven days, and like every web server's it records the address a request came from. We never ask it for that. The only two things this site asks the log are a path and a date, so what comes back is “the front page was fetched forty-one times on Tuesday” and never anything about a person — and because it is an aggregate of requests, it cannot tell whether anybody pressed play, or heard the end, or was the same somebody as yesterday. Nothing on this site counts a listen.
Comment authors from a scraped thread are anonymised before anything reaches a language model, so no name from a discussion is ever stored against an episode. Ask at the address above and the account and everything in it goes; it is done by hand, so allow a few days.