AI & Models
Earth AI aims to cut mineral sample processing from months to days
Earth AI is building its own labs, hoping to reduce mineral sample processing times from five months to five days to improve its AI exploration models.
Mineral exploration startup Earth AI is vertically integrating its operations by setting up its own laboratories to process rock samples. The company is hoping to reduce the turnaround time for processing these samples from five months down to just five days. By establishing these testing facilities, the startup aims to tighten the data feedback loop for its artificial intelligence models, which are deployed to locate critical minerals—defined as minerals essential for economic or national security.
The decision to build these facilities comes as the company faces severe data bottlenecks in its exploration projects in Australia. Typically, external laboratories that analyze rock samples have backlogs of around two months. However, Roman Teslyuk, the founder and CEO of Earth AI, noted that these industry delays have recently more than doubled. This surge in wait times has left the startup with a massive backlog of unprocessed material. “We’re 7 km behind — 7,000 meters of samples we don’t have data about.” Teslyuk said, explaining that the delays began to mount after the company increased its drilling capacity.
These prolonged wait times directly impact the efficiency of the exploration process. Because geological teams cannot visually confirm the presence of target minerals in drill cores—the cylindrical samples of rock extracted from the ground—they rely entirely on laboratory data to understand what lies beneath the surface. Teslyuk explained that waiting five months for results means the subsequent decisions on where to position the drill are not as effective as they could be. To minimize drilling, the company needs to get information in time to narrow down exactly where to go next, ensuring the drill is sent to the right spots to obtain the best data for the model.
While the internal laboratories will handle the rapid testing required to guide active drilling operations, Earth AI will still use third parties to validate its mineral discoveries before making final economic decisions. This external validation remains necessary for final decisions on the economic value of a mine, including those that might influence a sale.
Why it matters
By building its own testing facilities, Earth AI bypasses the severe backlogs of commercial laboratories, tightening the feedback loop for its exploration models. This vertical integration could significantly accelerate the discovery of critical minerals by allowing exploration teams to make rapid, data-driven drilling decisions.