Mendral offers an AI‑driven DevOps Engineer that automates security, reliability, and performance tasks for software teams. Its always‑on agents continuously review dependency pull requests, diagnose and fix flaky CI failures, and optimize build pipelines, while allowing custom automations triggered by CI/CD events or infrastructure alerts. The solution installs as a GitHub App in minutes and learns from each intervention to improve future recommendations.
Funding
Funding not disclosed
Founders
Product
Problem
Engineering teams spend significant manual effort on repetitive DevOps tasks such as reviewing dependency updates, triaging flaky CI tests, diagnosing build failures, and optimizing build performance. This overhead reduces development velocity and increases the risk of security and reliability issues slipping into production.
Solution
Mendral provides an AI‑driven DevOps Engineer that operates as three always‑on agents for security, reliability, and performance. The Security Agent continuously reviews dependency pull requests, pins safe versions, and surfaces only exploitable CVEs while detecting compromised packages and leaked secrets. The Reliability Agent automatically diagnoses CI failures, fixes flaky tests, and opens corrective pull requests. The Performance Agent identifies caching opportunities, parallelism gains, and slow tests to prune, reducing build times. Users can also define custom automations triggered by CI/CD events, infrastructure alerts, or external tools such as Slack or Linear. The system installs as a GitHub App in minutes, reads CI logs immediately, and learns from each investigation to improve future actions.
Target Audience
Primary customers are software engineering organizations that run CI/CD pipelines—particularly teams using GitHub Actions who need automated security, reliability, and performance improvements at scale.
Features
- Security Agent that reviews dependency PRs, pins safe versions, and highlights only CVEs exploitable in the codebase
- Reliability Agent that diagnoses CI failures, auto‑fixes flaky tests, and creates remediation pull requests
- Performance Agent that discovers caching wins, parallelism opportunities, and removes slow tests to accelerate builds
- Custom automation framework allowing triggers from CI/CD pipelines, infra alerts (Datadog, Sentry), schedules, Slack, Linear, or webhooks
- GitHub App installation in under five minutes with immediate CI log ingestion and real‑time analysis
- Continuous learning loop where accepted fixes and feedback refine future investigations and suggestions
- Integration with SOC 2 Type II compliant infrastructure and support for GitHub Actions (with Buildkite, CircleCI, GitLab CI forthcoming)