Changelog & Blog

Release notes, feature announcements, and technical deep-dives.

AllReleaseDeep DiveAnnouncement
Mar 25, 2026v3.2.0Release

Parallel Matrix Builds & Build Insights Dashboard

Run test stages across multiple OS/runtime combinations simultaneously. Plus, a new analytics dashboard to identify bottlenecks and flaky tests.

  • Matrix builds support up to 50 parallel jobs
  • Build insights with duration trending and cost attribution
  • Auto-quarantine for flaky tests (>3 flakes/week)
  • Improved cache invalidation accuracy by 12%
Feb 18, 2026v3.1.0Release

Native Container Registry with Vulnerability Scanning

Built-in OCI-compliant container registry with Trivy-powered scanning, Cosign signing, and SBOM generation. No more juggling external registries.

  • Push images directly from pipeline stages
  • Automatic vulnerability scanning (critical/high/medium/low)
  • Cosign + Sigstore image signing
  • SPDX and CycloneDX SBOM generation
Jan 10, 2026v3.0.0Major Release

PipeForge v3: Complete Platform Rewrite

Rebuilt from the ground up with a Rust-based scheduler, Go API gateway, and completely redesigned pipeline YAML syntax. 3x faster than v2.

  • New Rust-based DAG scheduler (3x throughput)
  • Simplified YAML syntax with smart defaults
  • Go API gateway with gRPC internals
  • New CLI with local execution mode
  • Backward-compatible migration tool included
Dec 05, 2025Deep Dive

How We Built a Content-Addressable Build Cache

A technical deep-dive into PipeForge's caching layer — from Merkle trees for dependency hashing to distributed cache nodes with 99.7% hit rates.

  • Content-addressable storage using BLAKE3 hashing
  • Distributed cache nodes with consistent hashing
  • Branch-aware cache inheritance
  • Automatic garbage collection with LRU eviction
Nov 12, 2025Deep Dive

Scaling to 2 Million Builds Per Day

The architecture behind PipeForge's build infrastructure. How we handle peak loads, ensure isolation, and maintain sub-second scheduling latency.

  • Kubernetes-native worker pool with auto-scaling
  • Firecracker microVMs for build isolation
  • Redis Streams for job queue with exactly-once delivery
  • P99 scheduling latency: 180ms
Oct 20, 2025v2.8.0Release

Deployment Targets: Kubernetes, Serverless, and Edge

First-class deployment integrations with canary, blue-green, and rolling strategies. Deploy from your pipeline with zero external config.

  • 12 deployment targets (K8s, Lambda, Cloud Run, etc.)
  • Canary deployments with metric-based promotion
  • Automatic rollback on error rate / latency breach
  • Multi-region deployment support
Sep 08, 2025Announcement

PipeForge Enterprise: On-Premise & VPC Deployment

Bringing PipeForge to your infrastructure. SOC 2 Type II certified, HIPAA BAA available, with dedicated support engineering.

  • Deploy in your AWS, GCP, or Azure VPC
  • Air-gapped installation support
  • SOC 2 Type II + HIPAA compliant
  • Dedicated support engineer for every Enterprise customer
Aug 01, 2025Deep Dive

Designing a Pipeline DSL That Developers Actually Like

The design decisions behind PipeForge's YAML pipeline syntax. Why we chose YAML, how we avoid the pitfalls, and what makes it different.

  • Smart defaults eliminate 60% of boilerplate
  • JSON Schema validation with IDE integration
  • Expression language for dynamic values
  • Template system for reusable pipeline fragments
Load More Posts

Stay in the loop

Get release notes and engineering posts delivered to your inbox.

you@email.com
Subscribe