Transforming Fleet Management Software Sales

According to details from OpenAI News, Proaction designs software solutions for companies managing vehicle fleets ranging from cars and trucks to heavy construction machinery. Because every operational environment is distinct, demonstrating how the platform integrates with a prospective client's business is a critical component of the sales process. You can learn more about these developments directly via OpenAI News. Further details are available from OpenAI News in the original source material.

Previously, developing personalized demonstrations demanded engineering resources that the team could not spare, forcing the founders to rely primarily on general conversations and slide decks. The introduction of Codex shifted this approach entirely, allowing non-technical team members to directly configure interactive environments.

Customizing Demos and Driving Sales Growth

Following a sales call, Colin Knudsen—co-founder and COO of Proaction—directs Codex to review Granola recordings, prospect email threads, and shared spreadsheets. Codex leverages this context to construct an HTML demo environment that mirrors both Proaction's product and the customer's specific fleet setup.

When sharing his screen, Knudsen presents prospects with visualizations of their own vehicles and equipment categorized by operational workflow. Knudsen estimates that the share of deals transitioning from initial contact into solution development rather than simple nurture phases has risen by 50% to 60% as a result of these customized demonstrations.

Saving Engineering and Administrative Hours

Producing comparable demonstrations internally used to take approximately 10 hours of engineering time per instance. With Knudsen now generating four to six customized demos monthly, the process saves between 40 and 60 hours of engineering effort every month.

In addition to reducing front-end sales barriers, the customized demo serves as a visual reference for engineers once a prospect converts into a paying customer. This continuity minimizes clarifying questions and reduces back-and-forth communication during production development.

Beyond customer demonstrations, Knudsen utilizes Codex plugins connected to tools such as Granola, Gmail, Slack, Linear, GitHub, and HubSpot to centralize customer context. By automating administrative tasks and pulling transcripts to draft follow-ups, Knudsen estimates he saves an additional 25 to 33 hours every month across 15 to 20 distinct daily responsibilities.

Integrating Voice Agents with GPT-Live-1 and GPT-6 Astra

Proaction also integrates OpenAI models deep into its core platform architecture. When users submit photographs alongside vehicle issue reports, ChatGPT-5.6 Sol assists in identifying equipment damage.

Furthermore, Proaction implements GPT-Live-1 and GPT-6 Astra to build specialized automated agents designed to handle day-to-day fleet operations. Termed the Managed Execution Layer, these agents execute phone calls, review documents and images, analyze text, and respond dynamically via chat.

An example of this capability is Marty, an agent engineered to coordinate vehicle maintenance. Marty communicates directly with drivers regarding reported issues, contacts repair shops, arranges service schedules, and manages estimate approvals before human teams step in for final review.

Accelerating Product Development

The implementation of GPT-6 Astra enables Proaction to scale these advanced customer experiences more rapidly. Danny O'Halloran, Head of Product at Proaction, noted that Astra's computer-use runs operate more succinctly compared to longer execution runs required previously.

Together, Codex and GPT-6 Astra provide Proaction's team with expanded capacity to sell, service clients, and advance the platform. Meanwhile, autonomous agents built on GPT-Live-1 return valuable time directly to fleet managers across the industry.

Sources

• OpenAI NewsProaction boosts sales 60% and saves 75+ hours with Codex