GitHub Next
GitHub Next investigates the future of software development. It is a team of researchers and engineers at GitHub that prototypes tools and technologies beyond the adjacent possible and explores new approaches to healthy, productive software engineering teams.
At a Glance
- Software developers and engineering teams
- GitHub repository maintainers and open-source communities
- Product managers, designers and other software-planning stakeholders
- Data scientists, journalists and other developer-adjacent data users
- +1 more
AI Tools by GitHub Next
(2)
LocalJev
Local Jev Compatible API Server
Copilot Workspace
AI Task Based Dev Environment
Discussions
No discussions yet
Be the first to start a discussion about GitHub Next
Latest News
Sparkle Pills: a flexible templating convention for contextual inference
Chopin published as a real-time multiplayer, agentic planning environment
Knowledge Compressor explores reducing documentation token counts while preserving usefulness
WikiKB introduced as a repository-native knowledge base for people and agents
Products & Services
A local, Jev-compatible POST /v1/systemone API written in TypeScript for Bun, backed by DiffusionGemma through an OpenAI-compatible Chat Completions endpoint. It translates typed Jev questions into classification prompts, validates and normalizes model outputs, and returns Jev-compatible decisions.
A real-time multiplayer, agentic planning environment intended to make team planning with agents collaborative, reduce cognitive load, and support decision-making and research.
A goal-driven Agentic Workflow that lets users define a goal, target and evaluation metric, then iteratively proposes and keeps improvements in a repository.
A research demonstrator for natural-language programming over GitHub Actions. Natural-language Markdown is compiled into executable Actions workflows that run with an agentic engine.
Market Position
GitHub Next is differentiated as an internal GitHub research-and-prototyping group, not a conventional standalone software vendor. It explores future developer workflows at the intersection of GitHub, GitHub Actions, Copilot, repositories and AI agents; relevant adjacent alternatives include commercial AI coding environments such as GitHub Copilot, Cursor, Claude Code and OpenAI Codex, workflow automation platforms, and repository-knowledge tools, but Next's stated role is to investigate and prototype rather than compete through a single product.
Leadership
Executive Team
Idan Gazit
Senior Director of Research
Early GitHub Next member who helped bootstrap the group from the Office of the CTO; previously a principal engineer at Heroku and an alumnus of the Django core development team.
David Slater
Director of Research
Joined GitHub in December 2022 working on Copilot modeling and evaluation benchmarks and joined GitHub Next in May 2024; previously principal investigator on DARPA research programs at Two Six Technologies.
Founding Story
GitHub Next grew out of GitHub's Office of the CTO; Idan Gazit describes himself as an early member who helped bootstrap the group from its beginnings. The initial vision was to explore how software development should work and to prototype new tools openly, rather than operate as a conventional product company.
Business Model
Revenue Model
GitHub Next is an internal GitHub research and prototyping team rather than a separately marketed, standalone revenue business. Its projects are generally research prototypes, open-source tools or experiments; some explorations can feed GitHub products. Public pages do not present a standalone GitHub Next subscription or API pricing model.
Target Markets
- Software developers and engineering teams
- GitHub repository maintainers and open-source communities
- Product managers, designers and other software-planning stakeholders
- Data scientists, journalists and other developer-adjacent data users
- Organizations experimenting with AI-assisted software engineering
- Planning and reviewing software with teams and coding agents
- Automating repository maintenance, CI repair and pull-request readiness
- Code migration, optimization and test-coverage improvement
- Natural-language automation of GitHub repository workflows
- Creating lightweight data-driven applications from versioned repository data
- Building and querying persistent project knowledge for humans and agents