DevRev delivers autonomous self-service with GPT-6.1 Sol
TL;DR: OpenAI published a customer story describing how DevRev uses OpenAI models, including GPT-6.1 Sol and GPT-6 Astra, to power its Computer and Operator agents for enterprise self-service. The story matters because it frames autonomous service agents as a systems problem: model reasoning must be paired with company context, permissions, reversible actions, source-backed answers, and acceptable latency and cost.
What DevRev built with OpenAI models
DevRev built Computer and Operator to help service and operations teams carry difficult customer and employee requests through to completion. According to DevRev Chief Technology Officer Ahmed Bashir, Computer and Operator are designed to reason across live data, take reversible actions, and point answers back to their sources.
DevRev combines OpenAI models, including GPT-6 Astra and GPT-6.1 Sol, with its own harness for connecting company knowledge and giving teams control over the actions an agent can take. OpenAI models power the reasoning, planning, and tool selection behind both Computer and Operator.
Why DevRev emphasizes reasoning, trust, latency, and cost
DevRev evaluates model choice as a balance between reasoning quality, reliability, alignment, latency, and the cost of serving customers and employees. Ahmed Bashir describes the requirement as a combination of capabilities rather than a single benchmark result.
“It’s a combination of being able to have the right reasoning capabilities, being able to be trusted, highly aligned, and most importantly, it has to work from a latency perspective.”
—Ahmed Bashir, Chief Technology Officer, DevRev
DevRev says GPT-6.1 Sol performs well in its evaluations. Bashir says the model needs to understand the nuances of a customer or employee request and respond quickly at a sustainable cost. “I think it strikes the balance perfectly,” he says.
Company context, permissions, and Enterprise-Bench
DevRev’s Agent Studio lets users describe how they want a tool to behave in a conversation with AI and define parts of its behavior with fixed rules. This gives teams a way to combine conversational instructions with deterministic controls.
DevRev tests OpenAI models together with the software that supplies company context, enforces permissions, and controls actions. Its Enterprise-Bench benchmark evaluates how well the models and surrounding system work when data has multiple versions, information sits in different systems, and access depends on a person’s role. DevRev says GPT-6.1 Sol performs well in those evaluations.
Codex adoption inside DevRev engineering
DevRev’s engineering team uses Codex for software development. Bashir says adoption grew organically after engineers explored multiple coding tools. By the time of the interview, DevRev had moved fully to Codex and dropped its Cursor licenses.
“Developers are really just looking for the best tool, and you can’t fool them.”
—Ahmed Bashir, Chief Technology Officer, DevRev
Bashir says DevRev engineers were unequivocal about wanting to continue with Codex.
“We think that Codex is far and away the best experience.”
—Ahmed Bashir, Chief Technology Officer, DevRev
DevRev’s sales engineers also use GPT-6 Astra to create demos for the company’s booths at international conferences.
Faster resolution with the Decisions API and Ultrafast
When a customer or employee asks for help, the request may require Operator to understand the issue and take an approved action, or route the request to Computer so the right person can act. Bashir sees opportunities for the Decisions API and Ultrafast to help those workflows finish quickly without requiring deep reasoning at every step.
Some decisions are difficult and expensive to define with fixed rules, especially when the information they depend on keeps changing. Bashir wants builders to describe more of that logic conversationally and choose the right kind of intelligence for each task.
“These things have to be able to run and complete themselves in seconds, not minutes.”
—Ahmed Bashir, Chief Technology Officer, DevRev
The next ambition: longer-running, traceable self-service
DevRev’s next ambition is for Operator to take on longer and more demanding customer and employee requests while keeping actions traceable, permissioned, and reversible. Bashir describes familiar support tasks, such as finding a knowledge-base answer, upgrading a plan, or getting a refund, as only the beginning.
“We want to really realize that promise of self-service.”
—Ahmed Bashir, Chief Technology Officer, DevRev
With OpenAI’s frontier intelligence, DevRev is building toward Computer and Operator carrying more complex requests through to resolution, with the company knowledge and controls needed to act on a person’s behalf.