Why this document exists
Search used to be an edge. For twenty years, the investor who searched better invested better. Finding the obscure filing, the early thread, the primary source before anyone else was a real, repeatable advantage.
That era is over. AI gives everyone the same answer at the same time. The moment information becomes universally retrievable, it stops being differentiating. Meanwhile the volume of content has exploded: feeds, forums, influencers, on-chain dashboards, research notes, all competing for attention, most of it noise.
Retail investors carry the heaviest cost of this shift. They cannot follow every source across every language and time zone. They cannot separate signal from noise on anonymous boards. They cannot trust engagement-driven algorithms that optimize for outrage rather than accuracy. The information asymmetry did not disappear when search got smarter. It changed shape.
We believe the answer is not more information. It is curation: the disciplined work of deciding what matters, for whom, and why. This document describes how we think that work should be done. It is our method, and it is the standard we hold our own products to.
01
Curation is the new search
When everyone gets the same answer, the valuable question is no longer "what is true?" but "what should I be looking at today?"
Retrieval is solved. Judgment is not. The next decade of investing tools will not be won by whoever indexes more data, but by whoever compresses it best: taking the day's price action, on-chain movements, research, and community discussion, and reducing it to the few things a specific person actually needs to see.
Our benchmark is simple and strict: five minutes a day should be enough. If a reader needs thirty minutes to feel caught up, the curation has failed. Compression is the product.
02
Your portfolio is the query
Relevance is personal. Two investors reading the same feed have wildly different needs, because they hold different assets, watch different sectors, and carry different risks.
Most platforms resolve this with engagement algorithms: show people what they click. We think that is backwards. Clicks measure impulse, not need. What someone holds and what someone watches is a far more honest signal of what they need to know.
So we build systems where the user's own context drives the feed. Connected assets, declared interests, and observed patterns decide what surfaces and in what order. The feed is not a broadcast. It is an answer to a question the user never has to type: "given what I own, what changed today?"
03
Humans and AI curate together
Algorithms scale. Editors judge. Neither is sufficient alone.
Pure AI curation amplifies whatever the data over-represents: hype cycles, coordinated shilling, recycled takes. Pure editorial curation cannot keep pace with a market that moves around the clock in a dozen languages.
Our answer is a hybrid scoring system we call Editorial Boost. Every piece of content is weighed on multiple axes: an editor's assessment of its quality, the authority of its source and author, genuine engagement, and freshness. Machines do the filtering at scale; editors set the standards and catch what models miss. The result is a feed where a verified analyst's measured note can outrank a viral rumor, by design.
04
Signal over noise, always
Most market content is volume without value. Repetition dressed as confirmation. Speculation dressed as analysis.
We treat noise reduction as a core engineering problem, not an editorial afterthought. Claims are cross-verified across independent source types: social posts against on-chain data, news against primary documents, community sentiment against actual holder behavior. Content that cannot be corroborated is deprioritized or discarded. Signals that could plausibly move a portfolio are separated from those that merely fill a feed.
We would rather show a reader three things that matter than thirty things that might. Every item we surface is a claim on someone's attention, and we treat that claim as a cost we must justify.
05
Utility beats incentives
The dominant growth playbook in Web3 is to pay for attention: airdrops, points, reward pools. It works, briefly. Then the pool runs dry, the incentive disappears, and the users leave with it. Rewards buy visits. They do not buy users.
We refuse that trade. Retention has to be earned the only durable way: by being useful every single day. A reader should return because yesterday's briefing was accurate, because the alert that fired last week saved them from a bad entry, because the platform demonstrably understands their portfolio better than any generic feed.
This principle also governs how we integrate with other services. No forced sign-ups, no mandatory wallet connections, no dark patterns. Every connection a user makes should be a choice they made because the value was obvious. Adoption that has to be coerced is adoption that will not last.
06
Value should compound
A tool that helps one user once is good. A tool whose usefulness deepens as usage accumulates is a platform.
We design for compounding from day one. Each briefing read, each asset watched, each alert configured makes the next recommendation sharper for that user. And in aggregate, the behavior of many informed users becomes something larger: a live map of market attention and sentiment that no single data vendor can replicate.
This is the arc we build toward. What starts as a daily habit becomes a data asset. What becomes a data asset can become infrastructure: a trust and reputation layer for a market that badly needs one. The user's value is never sacrificed for the platform's; the platform's value is the accumulated exhaust of genuinely serving users.
This principle has a shape. See the arc →
07
Ship from day one
We are not discovering this domain. We have operated content platforms reaching millions of monthly users, run editorial teams through multiple market cycles, and shipped products across finance and AI.
That history changes how we build. We do not need a cold start, a content pipeline built from scratch, or a year of learning the market's rhythms on someone else's time. Our curation standards were forged in a live newsroom, and our engineering has been tested in shipped products and winning hackathon builds.
Method comes before roadmap. Our arc, from a daily tool to infrastructure, is where these principles lead when they are applied in order. The way a team decides what is true, what is relevant, and what deserves a user's five minutes: that is what endures. This document is that way, written down.
Oreum builds curation you can trust, in the age of broken search.