Know what
matters.
Reels, podcasts, newsletters, articles, links. Bearing reads them before you do and tells you which ones are worth your time.
Too much input, no real filter.
The problem is not the volume. The volume has been this bad for years. The problem is that nothing sits between the inflow and your knowledge and decides beforehand. So you decide afterwards, in a folder full of things you saved because they looked important.
No context before consumption
A 40-minute podcast looks exactly like a 40-minute podcast you actually needed. You find out at minute 40.
One person, several jobs
Your feed does not know that you moved from the product question to the market question ten minutes ago. It shows you the same things either way.
Platforms set your priorities
Feeds optimise for time spent, because that is their business model and not yours.
Context management has three corners.
Organising, informing, knowing: three different jobs that usually get sold as one product, and that product is almost always a place to put things. Bearing works the middle one of them.
Bearing decides what your knowledge system gets in the first place. You do that job yourself right now, every day, and mostly without noticing.
Organisation management
Where things have their place
What exists, where it sits, who owns it, what is free right now. The storing happens elsewhere, this is only the floor plan.
Information management
What is worth your attention
Everything arriving from outside, and one question about it: is this worth your next forty minutes? Answered before you spend them, and per profile.
Knowledge management
What you already know
Corporate LLMs, RAG stacks, a wiki, or a folder structure somebody invented on a Tuesday that everyone has used since. They all keep things well, but what goes in there is still your call.
From transient input to deliberate context.
Six steps from reel to briefing. Step five is the interesting one: that is where it gets decided what never gets stored at all.
Reels, podcasts, newsletters, RSS, links, uploads. Bearing starts at the real media input, not at a saved article.
Supadata, Groq Whisper
Relevant for whom? The same signal ranks differently for your product, private, market or team profile.
Embeddings, pgvector
Watch, discard or promote. What you dismiss counts as much as what you read, and both sharpen later rankings.
Reading List, feedback loop
Transcript, entities, source, freshness and format are resolved before anything gets ranked.
GLiNER, OCR, parsers
Only when a signal stays ambiguous. Then it gets checked against current sources, counter-sources and the Spaces you already have.
Live Research, web and API adapters
Daily, weekly or on demand, with relevance, contradictions and a clear so-what per profile.
Structured outputs, digests
Reels, podcasts, newsletters, RSS, links, uploads. Bearing starts at the real media input, not at a saved article.
Supadata, Groq Whisper
Transcript, entities, source, freshness and format are resolved before anything gets ranked.
GLiNER, OCR, parsers
Relevant for whom? The same signal ranks differently for your product, private, market or team profile.
Embeddings, pgvector
Only when a signal stays ambiguous. Then it gets checked against current sources, counter-sources and the Spaces you already have.
Live Research, web and API adapters
Watch, discard or promote. What you dismiss counts as much as what you read, and both sharpen later rankings.
Reading List, feedback loop
Daily, weekly or on demand, with relevance, contradictions and a clear so-what per profile.
Structured outputs, digests
Contexts with filters, not just folders.
A profile defines the lens, a Space defines the working context. Bearing separates two questions that otherwise collapse into one: who is looking, and at what.
Which is why a Space has its own filter, its own briefing cadence, and its own rules about what gets in.
- What is a Space
- A Space bundles profile, inputs, briefing cadence, and promotion rules around one working context: Product & AI, Founder Ops, or your private weekly radar.
- What is a Briefing
- A condensed decision layer with relevance reasons, contradictions, blindspots, and a clear so-what.
Sunday Briefing
SW labs, Product & AI
- 0.94
OpenAI and Anthropic shift price anchors for coding agents
5 sources, 2 contradict each other on monetization
- 0.61
New MCP patterns are converging across the agent ecosystem
4 sources, early convergence
- 0.28
Browser-native AI operators show up in niche workflows
3 sources, evidence still thin
Information state: Founder- and vendor-heavy around agent tooling
Validate with Live Research12 signals evaluated live
- Agent Economics0.44
- Platform Shifts0.33
- AI Policy0.23
See what colors your input. Before it becomes knowledge.
Not media bias in general. Yours. The Bias Network shows which sources, perspectives, and repetitions currently dominate your information state, and marks the gap inside the briefing itself.
- Before information bias becomes knowledge bias
- When the same perspectives keep repeating, Bearing shows it early. Before they settle into your knowledge systems as apparent truth.
- Position shows the perspective
- Every bar sits where that cluster output actually lands on the axis. That is how you see what would otherwise stay invisible.
- Width shows the weight
- A wide bar took a lot of your time lately. That is a decision, even when nobody made it.
Paid for by you, not by advertisers.
That is the whole reason Bearing is allowed to tell you something is not worth your time. A product paid for by advertisers could not afford that sentence.
Solo
One user. One context. Full pipeline.
- Up to 50 sources
- 3 Spaces
- Daily briefings
- Bias Network
- Live Research
Team
Deeper context. Richer connections.
- Up to 200 sources
- 10 Spaces
- Custom briefing cadence
- Profile connections
- Chat agent
- Shared Spaces
Enterprise
For organisations with advanced needs.
- Unlimited users
- Unlimited sources
- Tenant context layer
- Custom integrations
- Dedicated support
- Rollout guidance
Preliminary pricing
Nothing has been billed yet, and these numbers can still change before launch. If you are on the waitlist you hear about it beforehand.
What gets asked most.
Ten questions that come up before every conversation about Bearing. If yours is not here, write to me.
What is Bearing?
Bearing is a relevance layer between your incoming media and your knowledge system. It reads reels, podcasts, newsletters, articles and links before you do, scores every signal against your profiles, and condenses the result into a briefing. No feed, no timeline, no unread counter.
How is Bearing different from a feed reader or a read-later app?
A feed reader and a read-later app both collect what arrives and leave the deciding to you afterwards. Bearing decides beforehand. The difference is the one sentence neither of them will say: that something is not worth your time.
What is a Space?
A Space is a working context with its own filter. It bundles profile, sources, briefing cadence and the rules for what gets in — Product and AI, Founder Ops, or a private weekly radar, for instance. A profile defines the lens, a Space defines what you are looking at.
What is a Briefing?
A briefing is a short text you read in a few minutes. It says what was relevant and why, where sources contradict each other, and what follows from that. Daily, weekly or on demand, and different per profile.
What is the Bias Network?
The Bias Network shows which perspectives currently dominate your information state — not media bias in general, yours. Every source you read sits on an axis, and its width shows how much of your time it got. Where nothing reaches you, Bearing writes the gap into the briefing.
Does Bearing handle video and podcasts?
Yes. Reels, podcasts and talks get transcribed and read in full. The verdict is about what was said, not the title and not the thumbnail.
Is Bearing just a ChatGPT wrapper?
No. A wrapper passes your question to a model and hands back the answer, and next session the desk is empty again. Bearing depends on remembering: what you read and what you dismissed becomes the context the next signal gets judged against. Models are components inside it, the way a database is a component.
What does Bearing cost?
Solo is 19 euros a month, Team is 49 euros, Enterprise on request. These are preliminary prices: nothing has been billed yet, and they can still change before launch. Bearing is paid for by its users and not by advertisers, and that is exactly why it is allowed to tell you something is not worth your time.
When can I use Bearing?
Bearing is not public yet. There is a waitlist, and the people on it hear about prices and dates before anyone else. There is no launch date yet.
Who builds Bearing?
Bearing is built by Felix Winkel at SW labs. Product, design and code sit with one person, decisions are made quickly, and you talk to the person who builds it.
Better ask first.
Press, partnerships, or the question whether Bearing actually solves your specific problem. Sometimes the answer to the last one is no. You will hear that too.
“I saved things for years. Bookmarks, watch later, playlists, a folder system I was genuinely proud of. I went back to almost none of it. At some point it was obvious that filing was never the problem. What is missing sits before the filing: the thing that decides whether something should go in at all, and which of my subjects it belongs to. Everything I read turns into what I think. Bearing is an attempt to stop making that one decision myself at one in the morning. Give me two years.”