Overland AI: ground autonomy software and vehicle integration

Overland AI is a Seattle company building a business around off-road autonomy. Its range includes software for driving, a retrofit package for existing vehicles, an operator interface and vehicles manufactured under its own product name. This combination places it between an autonomy software supplier and a ground-system manufacturer, with commercial opportunities in both categories.

The distinction matters to a vehicle maker considering a partnership. Overland can help turn another manufacturer's vehicle into an autonomous product, but it also offers a complete platform that a customer might buy directly. Its public record increasingly includes production-related work alongside research programmes and demonstrations. The most useful way to understand the company is to separate those routes and identify which one each announcement actually concerns.

Four products, two principal sales models

The current product family comprises OverDrive autonomy software, the OverWatch interface, the SPARK retrofit package and ULTRA, Overland's own autonomous ground vehicle. The software and retrofit products address customers or partners that already have a vehicle. ULTRA gives the company a complete platform around which it can organise manufacturing, integration and support.

SPARK's March 2025 launch described combining OverDrive with the computing, sensing and vehicle-control components needed for a retrofit. At a commercial level, this bundles integration work that an established vehicle supplier might otherwise undertake separately. The claim that a package can be widely applied is a product proposition; a specific vehicle still needs its own agreed integration and acceptance scope.

OverWatch occupies another part of the relationship: the interface used to manage autonomous assets. The interface can influence training, support and how new products fit into an existing fleet. A vehicle sale therefore need not exhaust the commercial relationship. Software maintenance and updates may remain relevant throughout its life, although the public material does not disclose Overland's pricing or the proportion of sales associated with recurring software.

ULTRA changes the allocation of responsibility again. A buyer procuring an Overland-built vehicle has a different supplier relationship from one buying a retrofit for another manufacturer's platform. Vehicle availability, spares and physical maintenance become more directly connected to Overland's delivery capability.

Manufacturing is part of the company model

Overland announced the opening of its South Seattle factory on 20 February 2025. The company described the facility as supporting both its own platforms and the modification of commercial vehicles. Congressman Adam Smith attended the opening.

This is useful evidence of an industrial footprint beyond a software development office. It also connects the two principal sales models: the same organisation can prepare vehicles it designs and integrate its technology into vehicles sourced elsewhere. The announcement does not establish a verified annual production volume, so facility opening should be treated as capacity evidence rather than a delivery count.

For a component supplier, the resulting opportunity is more specific than simply selling into robotics. Parts and services may enter through an Overland vehicle programme, through a retrofit package or through engineering work on a partner platform. Each route can involve a different party approving the component and carrying the maintenance obligation.

From research evidence to a production agreement

DARPA's RACER programme page describes research focused on off-road autonomy algorithms rather than the development of a particular vehicle or sensor. It documents field experimentation and identifies Overland AI as a source of programme imagery. This provides public agency context for the company's research background.

Overland's February 2026 business announcement says it completed RACER in November 2025. That matters because the programme supplied a setting for repeated development and evaluation. It does not make every later Overland product a DARPA-approved commercial system. Programme participation, product development and customer acceptance answer different questions.

The following records show how the company's commercial position developed:

Date	Public milestone	What it establishes
February 2025	South Seattle factory opening	An in-house production and retrofit facility
March 2025	SPARK launch	A packaged route for adapting existing vehicles
January 2026	CAL FIRE evaluation announcement	A named civilian public-service application
February 2026	$100 million financing announcement	Capital for company expansion, including debt
June 2026	$20 million APFIT production announcement	A defined Marine Corps production agreement
August 2026	Burlingame office announcement	An additional engineering and robot-operations location

The 29 June 2026 announcement describes a $20 million production Other Transaction agreement through APFIT, working with the Marine Corps' Program Manager Ground Based Air Defense. The stated scope is production of autonomous ground vehicles in support of the Marine Air Defense Integrated System.

The release includes a statement attributed to the Marine Corps programme manager. That gives the announcement a named customer connection. The publicly reviewed material does not provide the agreement's full schedule, quantities or payment terms. Its importance is the move into a specified production relationship; the amount should remain an announced agreement value, not be represented as annual company revenue.

A civilian application with a different buyer

In January 2026, Overland reported an evaluation with the California Department of Forestry and Fire Protection at Camp Pendleton. Two ULTRA vehicles were used to demonstrate wildfire-support logistics and observation tasks. The announcement includes a statement attributed to CAL FIRE Deputy Chief Jack Worden.

This example broadens the company profile without requiring an assumption that a civilian market has already scaled. CAL FIRE was an evaluation partner in the described activity. The release does not announce a fleet purchase or a statewide service contract.

Commercially, the example is relevant because the underlying mobility problem also exists outside defense. An organisation may need supplies or equipment moved across difficult terrain while retaining oversight of the work. Procurement, liability and support arrangements would be shaped by the civilian operator's requirements, even where the underlying autonomy technology overlaps with military programmes.

Ownership, funding and employee visibility

Founded in 2022, Overland identifies Byron Boots as co-founder and chief executive and Stephanie Bonk as co-founder and president. Its 3 February 2026 announcement reported $100 million in new funding. The equity component was led by 8VC, with participation from existing and new investors.

The total explicitly includes a $20 million venture-debt facility from TriplePoint Capital. Calling the entire $100 million an equity round would misstate the company's financing. Neither the funding total nor the production agreement supplies an annual revenue figure. A current total employee count and annual turnover were not verified in the reviewed public sources.

Overland subsequently announced a Burlingame office on 12 August 2026, focused on software engineering and robot operations. It said it planned to double its Bay Area headcount over six months, but gave no absolute number. Seattle remained the headquarters identified in the announcement. The new office is evidence of geographic expansion, not a quantified workforce increase already achieved.

How to compare its position

BDI's ARX Robotics profile provides another view of the relationship between ground vehicles and autonomy software. The Shield AI profile offers a comparison with an autonomy company operating across software and its own platforms. These comparisons help clarify business models without suggesting that the products share the same customer requirements.

For Overland, the next commercially revealing evidence would connect production awards to delivered vehicles, support commitments and repeat purchases. Its retrofit business also merits separate attention: adoption by a vehicle manufacturer could create a different distribution path from direct government sales. The company has public evidence supporting both an integration role and a manufacturing role; the balance between them remains an important part of its developing business.