NEW AI Audit and Voice AI now sit alongside the platform — one login, one system. Explore Voice AI →
, ,

DevRev Builds Autonomous Self-Service on OpenAI’s GPT-6.1 Sol

DevRev uses GPT-6.1 Sol, a model from OpenAI, to deliver autonomous self-service. OpenAI published this on its own website. The post links a specific OpenAI model to an enterprise customer-service use case.

The news matters most to support leaders and operations teams. Many of them are deciding whether AI should only answer questions or also resolve requests from start to finish. A named deployment gives these teams a concrete reference point.

The verified details are limited to the companies involved, the model name and the self-service use case. Release status, pricing and performance figures are not covered here and should be checked directly with OpenAI and DevRev.

What OpenAI published about DevRev and GPT-6.1 Sol

The post on OpenAI’s site describes DevRev delivering autonomous self-service with GPT-6.1 Sol. It is framed around how one company applies the model, not around the model’s launch. Readers should treat it as a deployment example rather than a product announcement.

What autonomous self-service means

Self-service lets customers solve problems without waiting for a human support agent. Familiar examples include help centres, chat assistants and guided troubleshooting.

Calling it autonomous usually means the AI system does more than suggest an answer. It can work through a request across several steps, which may include retrieving information, choosing what to do next and completing a task. If any term here is unfamiliar, the AiLunaPro AI glossary defines the vocabulary used in agent-based automation.

That shift from answering to acting is the main change for support teams. It also raises the bar for oversight, because a wrong action costs more than a wrong suggestion.

Why the model choice matters in enterprise support

Teams choosing a model for customer-facing automation usually weigh three things:

  • Reasoning quality: whether the model can follow multi-step requests and company policies without drifting.
  • Latency: how quickly it responds, which shapes whether a conversation feels natural.
  • Operating cost: what each resolved request costs at scale, since support volumes can be large.

A public deployment with a named model shows which option one vendor picked. It does not tell you how that model would perform on your own tickets, policies and systems.

Practical implications for businesses

For companies already running AI in support, this kind of example is a prompt to review where automation stops today. Common boundaries include answering only, answering and then handing off to a human, or resolving the request fully.

For companies just starting out, the useful lesson is about sequencing rather than any particular vendor. Map which requests are repetitive and low-risk before you automate anything. A broader view of how agentic systems change operations is covered in our overview of agentic AI and the age of automation.

Open questions to raise with vendors

Before adopting autonomous self-service, ask any provider these questions:

  • Which request types does the system resolve end to end, and which does it escalate?
  • Which actions need human approval, such as refunds, account changes or data deletion?
  • How are model outputs logged, reviewed and corrected?
  • How is resolution quality measured, and over what time period?
  • What happens to customer data sent to the model provider?

How to evaluate autonomous self-service for your team

  1. Inventory your requests. Sort recent tickets by volume, risk and how complex they are to resolve.
  2. Define action boundaries. List what the AI may do on its own and what always needs a human approval point.
  3. Pilot on low-risk flows. Start with information lookups or status checks before moving to account changes.
  4. Measure against a baseline. Compare resolution rates, escalations and customer feedback with your current process.
  5. Review compliance early. Check data handling and regulatory obligations before you scale.

If you want an outside view of where automation fits your support operations, an AI audit of your processes can show which workflows are ready and which need safeguards first. Teams planning a dedicated agent can also read our guide to building custom AI agents for business functions.

Last updated: this article reflects the information OpenAI had published about DevRev and GPT-6.1 Sol at the time of writing.

Put this AI capability into practice

Understanding the technology is only the first step. AiLunaPro can help your team turn this capability into a controlled operational workflow.

With AiLunaPro All-in-One AI Platform, you can:

  • Helps support teams separate confirmed deployment details from vendor marketing when assessing AI models
  • Gives a practical checklist for setting action boundaries and human approval points before automating customer service
  • Connects model-selection news to concrete steps for evaluating automation in your own support workflows

Explore the AiLunaPro Platform

For a workflow designed around your organisation, discuss a custom AI agent with our team.

Sources

Verified on DANG! Fazier badge