dScribe AI's Autonomous Drones Measure the Grain Pile and the Gravel Heap

The YC-backed startup is selling 3D vision to replace manual surveys that can be off by 30% in agriculture and mining.

About dScribe AI

Published

A grain pile is not a pallet. A gravel heap is not a box. The bulk materials that underpin agriculture and mining are notoriously difficult to measure, often forcing operations managers to rely on rough estimates or expensive, infrequent manual surveys. dScribe AI, a Y Combinator S25 startup based in Olathe, Kansas, is betting that autonomous drones and fixed 3D sensors can replace the guesswork with precise, continuous measurement [Prospeo, April 2025].

The wedge in bulk inventory

The company's wedge is simple: materials like grain, sand, and gravel don't fit into neat, barcoded units. Traditional measurement methods can be off by 20 to 30 percent, according to the company [Prospeo, April 2025]. That error translates directly into financial loss, logistical headaches, and inaccurate forecasting for consumption and deliveries. dScribe AI's product suite uses computer vision and LiDAR to generate accurate 3D reconstructions of stockpiles, calculating volume and weight from autonomous site scans [Y Combinator / LinkedIn, May 2025].

They offer a tiered approach to data capture, a practical nod to the varied needs and budgets of industrial sites.

  • Fixed monitoring. For high-throughput sites, the company installs permanent cameras and 3D sensors tailored to the location for continuous tracking [Rebulk].
  • On-demand capture. For sites that don't justify fixed hardware, operations teams can use smartphones or drones to capture scans when needed [Rebulk].
  • Actionable insights. The core promise is to move beyond a simple volume number. The system aims to provide real-time tracking of material movements and reliable usage forecasts, giving operations teams a data-driven view of their inventory [dScribe AI].

The team and the YC stamp

The founding trio brings a mix of product development and domain-specific insight. Co-founders Cole Robertson, Jordan Mryyan, and Warren Wang have built and scaled products at companies like Microsoft, Garmin, and Tropic, according to their Y Combinator profile [Y Combinator / LinkedIn, May 2025]. Warren Wang, serving as COO, saw the inventory tracking problem firsthand as a Product Manager at a Series A AgTech startup, working directly with feed suppliers and transportation managers [Y Combinator / LinkedIn, May 2025].

Founder Role Key Background
Jordan Mryyan Co-founder & CEO Previously Software Engineer II at Microsoft
Warren Wang Co-founder & COO Former Product Manager at a Series A AgTech startup
Cole Robertson Co-founder & CTO Product development background at companies including Garmin

The $1.7 million seed round

In April 2025, dScribe AI closed a $1.7 million seed round [Prospeo, April 2025]. The investor syndicate is a mix of coastal tech funds and regional specialists. Y Combinator, Abstraction Capital, and Flyover Capital were joined by Kansas City-focused funds KCRise Fund and Redbud VC. Notably, EquipmentShare, a construction technology company, also participated [Prospeo, April 2025].

Where the wheels could come off

The bet is clear, but the path is not without friction. dScribe AI is entering a space with established players. The differentiation for dScribe AI must rest on the specificity of its bulk inventory algorithms and the operational workflows it builds for grain silos and aggregate yards, not just topographic mapping. The hardware-plus-software model also introduces complexity. Selling and deploying physical sensors and drones requires a different motion than pure SaaS. The real test will be whether the value of continuous, fixed monitoring justifies the capex for their target customers.

The next twelve months

For dScribe AI, the immediate milestones are tangible. They need to convert their seed funding into a shipped product and their first named customer deployments. The partnership with EquipmentShare could provide a valuable channel or pilot site [Prospeo, April 2025]. Landing a public reference customer in either agriculture or mining would be the strongest possible traction signal, moving the narrative from technical potential to commercial proof.

Read on Startuply.vc