Skip to content

Honest about where we are.

Every capability, in one of three states. We don't publish dates we can't promise. We do publish what's real.

Live Running in production today.

Delivery insight in minutes

Install the app and get a workspace with 90 days of history, kept current by webhooks and a nightly self-healing sync.

  • Sign in with GitHub, workspace created from a verified install
  • 90-day backfill that respects rate limits, with a per-run estimate
  • Cycle time, review wait, PR size and throughput
  • Dashboards per contributor, team and custom group
  • Custom groups that can follow a team automatically
  • Nightly reconcile and hourly redelivery repair any missed webhook
  • Guided setup checklist, command palette and a global filter bar

Every source that matters

One consistent model across code hosts, trackers, incident tools and CI, so a number means the same thing wherever the work happened.

  • GitHub, GitLab and Bitbucket Cloud
  • Linear, Jira Cloud and Asana, with tickets linked to pull requests
  • PagerDuty, incident.io and Opsgenie for incident-based DORA
  • CI runs and jobs, GitLab pipelines and deployments
  • One person across providers, linked automatically or by hand
  • Contributor attributes from a CSV upload

AI and agent impact

What AI-assisted and agent-authored work actually ships, next to work with no AI signal, across the tools you connect.

  • AI-assisted PR classification from commit and PR signatures
  • AI-assisted vs. no AI signal on cycle time, size, review depth and reverts
  • Linked bug and incident rates for AI-assisted vs. no-AI-signal PRs
  • Claude Code, Codex CLI and Gemini CLI telemetry over OpenTelemetry By plan
  • Adoption, acceptance and spend by team and tool
  • Agent sessions linked to the pull requests they produced
  • AI spend by ticket and initiative
  • Adoption date and a return-on-investment view
  • Human intervention rate and autonomous run length per tool
  • A developer's own private AI view

DORA and benchmarks

Deployments and incidents linked to the pull requests behind them, so the four DORA keys come from real events, not surveys.

  • Deployment frequency and lead time for changes
  • Change failure rate and time to restore, from deploys or incidents
  • Industry benchmark bands with weekly targets
  • Benchmark history and per-team target overrides
  • Ticket cycle time and WIP per person as benchmarks

Reports for the questions you have

A report for each question a delivery review asks, and a builder for the ones nobody predicted.

  • PR Workflow with stage breakdown and percentiles
  • CI Analytics across workflows, runs and jobs
  • Bug Creation, Bug Resolution and Bug Workload
  • Predictability: done vs. committed, rollover and scope creep
  • Initiatives with progress, forecast and linked pull requests
  • Collaboration network and cross-team hand-offs
  • Review load per team and reviewer, in place of leaderboards
  • Lightweight team pulse surveys
  • Report builder with saved custom reports

Dashboards, alerts and digests

DevGlimpse comes to where your team already works.

  • Custom dashboards with a dashboard editor By plan
  • Saved views, share links and a read-only viewer role
  • Alert rules delivered to Slack and Microsoft Teams
  • Nightly anomaly detection with markers and an alert condition
  • Weekly team, personal and channel digests
  • Working hours, holidays and business-time durations
  • Workspace time zone and per-team targets

API, MCP and org structure

Your delivery data, reachable from your own tools and shaped like your company.

  • MCP server so your AI assistant can query DevGlimpse, read-only
  • Public REST API with an OpenAPI description By plan
  • Org hierarchy with director roll-ups
  • Attribute filters and group-by, with a minimum slice size

Privacy and control

Individual metrics are scoped per role on the server. Team numbers stay visible to everyone.

  • Per-role visibility: everyone, reporting line, my groups, or only myself
  • Minimum group size, so a small team never exposes one person
  • Pseudonymous mode for individual metrics By plan
  • Developer, manager and org admin roles with approvals
  • Append-only audit log with filters and CSV export
  • Workspace IP allowlist By plan
  • A key per workspace for credentials and exports, destroyed on deletion
  • Workspace and per-person export, per-person erasure
  • 30-day read-only window before any offboarding deletion
  • WCAG 2.1 AA accessibility baseline

Building now In active development now.

More AI vendors

The same adoption and spend picture for the assistants that do not speak OpenTelemetry.

  • Org-level Copilot usage adapter
  • Anthropic and OpenAI usage and cost
  • Cursor usage adapter

On the roadmap Designed and queued. Design partners help set the order.

Metric language and insights

Define your own metrics and let DevGlimpse point at what changed.

  • Custom metrics with a formula editor
  • Opt-in narrative insights
  • "Ask DevGlimpse" chat over your own metrics
  • Code churn and time allocation

Outcomes

Connect engineering effort to the work the business cares about.

  • Business KPIs linked to initiatives
  • Investment allocation across initiatives
  • AI rules of engagement with an optional GitHub check

Enterprise

The controls a security review asks for.

  • SSO with OIDC and SAML, SCIM provisioning
  • GitHub Enterprise Server as a code source
  • Warehouse export
  • Self-hosted package, Helm chart and on-prem agent
  • EU data residency
  • SOC 2 Type 2 and an external penetration test

Build it with us.

Design partners: free until 31 January 2027, then a locked founding rate. Book 20 minutes.

Free to start
The Free plan tracks up to 10 contributors with no time limit, while we build together.
Founding pricing, locked
After the free period, the Founding rate, locked for 24 months.
A direct line to the builders
A shared channel with the engineers writing DevGlimpse, not a ticket queue.
Shape what ships next
Integrations, metrics and reports: your needs set the order.