Frequently asked questions about Flux
Answers to all your Flux questions, all in one place.
What Flux is
Q: What is Flux?
It analyzes real code, commits, and pull requests, not ticket data or manual reports, to give engineering leaders ground-truth visibility into velocity, team dynamics, and code quality across their entire codebase.
Q: Who is Flux built for?
Flux is built for engineering leaders—VPs of Engineering, heads of engineering, directors, and managers—who need objective visibility into how work is actually happening across the codebase. It’s designed for GitHub Cloud organizations with 50 or more engineers: teams where a leader can no longer personally track everything and needs estate-wide intelligence to lead effectively.
Tech leads and senior ICs also use Flux to understand code health, hotspots, and team dynamics without adding reporting overhead.
Q: Who uses Flux day to day—leaders or engineers?
Flux is built for engineering leaders but used across the organization.
Engineering leaders (CTOs, VPs of Engineering, and directors) use it to understand velocity, focus, and risk across teams and repos, and to explain tradeoffs and progress in executive language. Managers and tech leads use it to see how teams actually work in the code, identify force multipliers, and manage technical debt at scale. Senior ICs use it to understand code health and hotspots in the areas they own, without additional reporting or ticket hygiene.
Q: What problems does Flux help me solve?
Flux gives you ground-truth visibility into how work actually happens in your codebase, not what tickets say, and not what status reports claim. It helps you act early, before delivery friction, hidden risk, or shadow work shows up in an incident or a missed deadline.
Specifically, Flux helps you:
- Separate feature delivery from maintenance, refactoring, and rework using objective, code-derived metrics (including DORA). not reported activity.
- Spot emerging risk in PR size, churn, and complexity before it surfaces in sprint reviews, incidents, or escalations.
- Expose shadow work (refactors, research spikes, configuration changes) that never makes it into tickets but consumes capacity and distorts planning.
- Connect outages, performance issues, and security incidents back to hotspots in the codebase and accumulated technical debt.
- Explain progress, tradeoffs, and risk to executives and the board using code-first evidence instead of anecdotal status updates.
How Flux compares to other tools
Q: How is Flux different from engineering intelligence platforms like Jellyfish or Swarmia?
Traditional engineering intelligence platforms are ticket-first. They rely on Jira and other human-updated systems to reconstruct what engineering did. Flux is code-first: it derives insights directly from commits, pull requests, and code structure, so you see real work and AI-driven change at code speed, without enforcing ticket hygiene or adding admin overhead.
The result is a fundamentally different quality of signal. Ticket-based tools show you what teams reported. Flux shows you what actually happened.
Q: How is Flux different from code quality tools like Snyk or CodeRabbit?
Snyk and CodeRabbit catch and fix issues in individual code changes. Flux sits above them, showing patterns across your entire codebase and teams: where risk is accumulating, how AI is changing delivery, and how quality work maps to features, maintenance, and technical debt.
Flux doesn’t replace your scanners or AI code review. It gives you the system-level view those tools can’t provide.
Q: We already use AI code review tools. Why do we need Flux?
AI code review tools operate at the change level, catching problems in individual pull requests. Flux operates at the system level, surfacing patterns in risk, architecture, and quality across all repos and teams.
It answers the leadership questions PR-level tools and ticket dashboards can’t: Where is AI helping or hurting velocity? Where is technical debt slowing delivery? Which teams are force multipliers? How do I explain this quarter’s progress to the board?
Q: Does Flux replace Jira or our ticketing system?
No. Flux complements Jira by showing what actually happened in the code, not just what tickets say. Jira remains your system of record for intent and status. Flux adds ground-truth visibility that ticket data can’t provide.
Leaders use Flux to validate plans, spot hidden work, and connect incidents and roadmap outcomes back to the codebase, while tickets stay in place for coordination and planning.
Q: Does Flux support GitLab, Bitbucket, or Azure DevOps?
Flux is purpose-built for GitHub Cloud organizations. If your engineering estate runs on GitHub Cloud with 50 or more engineers, Flux is designed for you. Support for additional source control platforms is not currently planned.
Q: Does Flux support DORA metrics?
Yes. Flux tracks DORA metrics (deployment frequency, lead time for changes, change failure rate, and time to restore) derived directly from code and PR activity, without requiring manual input or ticket instrumentation. Because Flux uses code-first data, DORA metrics in Flux reflect what actually happened, not what was reported.
Setup and integrations
Q: Is Flux an app we install? What does setup look like?
Flux is a cloud SaaS platform. There’s nothing to install on developer machines. To get started, you connect your GitHub Cloud repositories and configure which repos and teams you want analyzed. No changes to your existing ticketing or delivery processes are required.
Most customers connect their repositories and start seeing classified data within minutes. Flux backfills historical commits and PRs automatically, so you’re not starting from a blank slate.
Q: Do we need to install agents or change our CI/CD setup?
No. Flux doesn’t require agents on developer machines or changes to your CI/CD pipeline. You connect your GitHub Cloud repositories and identity provider; Flux handles analysis from its own infrastructure. Setup overhead for your engineering team is minimal.
Q: Does Flux support monorepos and multi-repo estates?
Flux is built for estate-wide analysis and supports monorepos, multi-repo estates, and selective inclusion or exclusion of specific repositories. It’s designed to scale across hundreds or thousands of repositories, including large, complex environments where code volume and change velocity are high.
Q: Can we control which repos and teams are visible to which users?
Yes. Flux’s role-based access model lets you control which repositories are connected and what each user can see, so you can scope views by team and align access with your org structure and privacy requirements.
Q: Can I try Flux before connecting our GitHub repositories?
Yes. The Flux sandbox is pre-populated with real open-source data and requires no GitHub login or repository connection. You can explore the full product experience, including work classification, leadership views, and risk signals, before touching your own data.
Q: How long does it take to get value from Flux?
Once your repositories are connected and users authenticated, Flux automatically backfills historical pull requests and commits and begins generating insights. No ticket hygiene, no new processes, no waiting for data to accumulate. Most customers see meaningful code-level signals as soon as initial indexing completes.
Security, privacy, and compliance
Q: How does Flux keep our code safe and secure?
Flux runs on Google Cloud Platform with encryption in transit and at rest, per-customer data isolation, strict access controls, and audit logging. Your source code is treated as critical IP. It’s used only to generate your insights and support your team, and is never sold or shared with third parties beyond essential infrastructure providers like Google Cloud Platform.
Q: Do you have SOC 2 or other security certifications?
Flux is designed with SOC 2 and ISO 27001 requirements in mind. The security architecture and controls already follow those frameworks; formal certification is actively in progress as the platform scales.
Q: Do you train AI models on our code?
No. Your code is not used to train shared or third-party LLMs. Flux uses Gemini via Google Cloud’s Vertex AI, and Google’s data governance terms prevent customer data from being used to train shared models.
Flux may use your data to improve work classification and models strictly within your tenant. Insights and models are never shared across customers.
Q: Which LLMs does Flux use?
Flux uses Gemini via Google Cloud Platform’s Vertex AI. Customer code, prompts, and derived metadata are sent to the model only as needed to generate your insights. The underlying model is not fine-tuned on your data.
Q: Can we bring our own LLM?
BYO-LLM support is on the roadmap. The current architecture is built around secure, server-side LLM usage via Google Cloud Platform’s data governance model. Any model integration will maintain strict tenant isolation and enterprise-grade data governance.
Value, outcomes, and ROI
Q: What outcomes should I expect in the first 90 days?
After connecting your repositories, Flux backfills historical commits and pull requests automatically, so you see meaningful signals quickly without waiting for data to accumulate. In the first 90 days, most customers see:
- A clear work-type breakdown. Feature delivery vs. maintenance, refactoring, and technical debt, derived from code rather than ticket labels.
- Early velocity warning signals. Shifts in PR patterns, churn, and complexity that surface friction before it becomes an incident or a missed deadline.
- Executive-ready evidence. Code-first explanations of progress, tradeoffs, and risk to replace anecdotal status updates in board and exec conversations.
- A prioritized risk view. The highest-impact areas in your codebase to target for refactors and debt paydown.
Q: Why do I need Flux now that my teams are using AI to write code?
AI has increased the speed, volume, and complexity of code changes, but it’s also made visibility into that code harder and less reliable. Tickets capture even less of the real work than before. PR-level AI reviewers don’t show systemic patterns in risk, architecture, or technical debt across your estate.
Flux is built specifically for AI-accelerated engineering. It gives you code-first visibility at the same speed your AI-assisted teams are shipping, so you can see how AI is actually changing velocity, quality, and team dynamics, and bring that evidence into executive conversations.
Q: How accurate are Flux’s insights, and can I correct them?
Flux derives signals directly from code, commits, and pull request metadata, which eliminates the human bias that skews manual reporting. The platform is designed to incorporate your feedback and refine categorization (for example, feature vs. refactor vs. maintenance) over time, always within your tenant boundary.
The more you use Flux, the more precisely it reflects how your organization actually works.
Q: How is Flux priced?
Pricing reflects engineering org scale and repository footprint, not per-ticket volume or per-seat counts. For details and to find the right fit for your team, talk to us.