Atlas is live, and we are keeping the beta tag while we improve four parts of the product: the knowledge it draws on, the depth of its news history, its scoring, and the way it sorts results.

Those limitations affect what you see today. Tags have room to improve. A company's news history only reaches back to launch. Scores and the order of stories will evolve as we build a stronger foundation underneath them. Here is what we are working on, in priority order, and what needs to change before we remove the beta tag.

1. Bringing more knowledge into Atlas

Our first priority is getting more of our mining knowledge into the system so it can inform how Atlas interprets and tags information. We are testing a scoped knowledge system: a way to bring the relevant knowledge into the context where it is needed. Once testing is complete, we plan to deploy it to production. We expect this to improve tagging substantially. For users, the intended change is more useful labels on stories and a stronger basis for finding information relevant to their research. That improvement is still ahead of us. The system is being tested, and the beta tag reflects the gap between the knowledge we want Atlas to use and what is available to it today. Adding that knowledge is the first step because later improvements depend on it.

2. Building a deeper news history

Atlas currently contains news collected since launch. That gives company pages a short history, which limits how much context you can find when you start researching a business.

A small number of stories in Atlas should therefore be read in light of that limited collection window. It does not tell you how much a company has disclosed over its lifetime. We are iterating quickly on ingestion: the process that brings news into Atlas. Once that work is ready, we plan to backfill roughly a year of news. The goal is to give company pages more stories and a longer record to explore. That should make it easier to look beyond the latest announcement and follow earlier developments. Roughly a year is the intended backfill window; it is not a claim that Atlas already has a complete archive for every company.

3. Improving scores, then the broader views

More knowledge and a deeper news history give us a better foundation for improving the scores attached to individual stories. That is the next priority. We want scoring to make better use of the context available to Atlas, so a story can be assessed with more relevant information behind it. After improving those individual scores, we will work on how they are brought together across regions and commodities. These aggregate views need attention of their own; improving a story score is only part of making the wider picture useful. For users, this means both story-level scores and broader summaries will evolve. The sequence matters: first strengthen the information going in, then improve individual assessments, then improve how those assessments are combined.

4. Making sorting more useful

Our current sorting is basic. Once scoring improves, we will build a better sorting algorithm, starting with those improved scores. The order of results matters because it shapes what people notice first. A useful story can be hard to find if the ordering does not surface it at the right point in someone's research. Over time, we want to move towards sorting that better reflects users' preferences. Think of how an e-commerce store works to show customers products relevant to their interests. Atlas needs to keep getting better at helping people find information relevant to theirs. That is a longer-term direction, and it will continue after beta. The beta milestone is the initial improvement to sorting based on better scores. Preference-based ranking is not a requirement we are adding to that milestone.

What we are building alongside these improvements

Over the next one to two months, the only additional features we plan to add are:

Atlas Wiki.
Company Spotlights.
MCP integration.
A content management system, or CMS, for our marketing team to add content to Atlas.

These features will be worked on alongside the four core improvements. They are separate from the conditions for removing the beta tag. The one-to-two-month window describes our current planning horizon, rather than a promised date for completing beta.

What closes beta

These four improvements define the path out of beta: more complete knowledge ingestion, deeper news history, better story scores and regional and commodity summaries, and an initial improvement to score-based sorting. Once those improvements are in place, we can remove the beta tag. Sorting will remain an ongoing priority beyond that point. You can explore Atlas today today with these limitations in mind. This is the work we are focused on to make that research experience more useful.

Why Atlas, Our Mining Stocks Tool, is Still in Beta (And What's Next)