Table of Contents
Ridges, the team behind Bittensor subnet 62, has launched Ridgeline, a new AI-powered engineering agent designed to autonomously complete software development tasks such as GitHub issues.
The product is now available in open beta, allowing developers to assign engineering tickets to an AI agent that can analyze repositories, generate code patches, and test its own work without human supervision.
Ridgeline serves as the product layer built on top of the Ridges subnet, where decentralized AI miners compete to continuously improve the agent’s capabilities.
Access the Beta
An AI Engineer That Works From GitHub Issues
Ridgeline is designed to function like an additional member of a development team.
A user can assign a GitHub issue to the agent, which then reads the entire repository to understand its structure and dependencies. From there, the system generates a solution, produces the relevant code patches, and runs tests to validate the fix.
If tests fail, the agent diagnoses the problem and iterates until the issue is resolved.
The system allows developers to delegate engineering tasks to an autonomous agent that can work in parallel with human contributors.
How Bittensor Powers the System
Behind Ridgeline is the Ridges subnet on Bittensor, where miners compete to improve the underlying AI engineering agent.
Instead of relying on a centralized development team, the subnet uses Bittensor’s incentive model to reward participants who produce better results on specific software engineering tasks. These tasks can include writing tests, fixing regressions, or generating code patches.
Agent performance is continuously evaluated based on metrics such as correctness, speed, and cost, and the best-performing contributors earn rewards.
This competitive model allows the system to improve continuously as miners compete to deliver stronger agent performance.
Inference for Ridgeline runs across infrastructure connected to the Bittensor ecosystem, including compute providers such as Targon and Chutes, which handle the compute required for model execution.
A Product Layer for Decentralized AI
Rather than operating purely as AI research infrastructure, the Ridges team is positioning the subnet as a production system capable of delivering real software engineering output.
The team says launching a real product introduces a feedback loop between users and the decentralized AI system.
Developers assign real engineering tasks, and the subnet receives useful data about what problems need solving and how well the agents perform. That information can then drive improvements to the models competing within the subnet.
The approach also highlights a potential path toward revenue-generating applications within the Bittensor ecosystem, where decentralized AI infrastructure powers commercial software tools.
Open Beta Now Live
Ridgeline is currently available in open beta, with new users receiving 10 free credits to test the platform.
During the beta period, developers can experiment with assigning issues and evaluating how the autonomous agent performs on real repositories.
The launch positions Ridgeline as one of the first developer-focused products built directly on Bittensor’s decentralized AI infrastructure and offers a view into how competitive AI subnets may translate into practical tools for software teams.
Get Started
Disclaimer: This article is for informational purposes only and does not constitute financial, investment, or trading advice. The information provided should not be interpreted as an endorsement of any digital asset, security, or investment strategy. Readers should conduct their own research and consult with a licensed financial professional before making any investment decisions. The publisher and its contributors are not responsible for any losses that may arise from reliance on the information presented.
