Tech Radar

cpp.ai’s Tech Radar shares our current view of emerging languages, frameworks, tools, platforms, and practices. It guides client decisions with actionable, evidence‑based recommendations.

Categories

  • Adopt: Mature enough for production; we recommend by default.
  • Trial: Promising pilot with guardrails and success metrics.
  • Assess: Watch closely, run spikes; not yet client‑default.
  • Hold: Avoid for now; specific risks outweigh value.

Examples

Adopt

C++20/23 • CMake • Bazel • clang‑tidy • GoogleTest/Catch2 • ROS 2 (industrial) • gRPC/Protobuf • OpenAPI • Docker • GitHub/GitLab CI • Cantata (testing) • Klocwork (static analysis) • Lattix (architecture) • SPDX/CycloneDX SBOM • Sigstore (signing) • Zephyr RTOS (selected use cases)

Trial

Rust for safety‑adjacent components • RISC‑V on embedded targets • libFuzzer/AFL++ • CodeQL • eBPF observability in regulated environments • Bazel remote caching/execution • Hardware‑in‑the‑loop at scale (farm patterns) • SLSA Level 3 provenance • Model‑based design bridges to traceability

Assess

Zig for systems glue • Wasm/WASI for sandboxed plugins • GPU offload on edge (Jetson Orin) for perception • Formal methods toolchains integrated into CI • High‑assurance containers on real‑time OSes

Hold

Unvetted AI code generation inside safety‑critical code paths • Ad‑hoc vendor forks without upstream strategy • Large monorepos without trunk‑based discipline or automated quality gates • Out‑of‑tree kernel patching for long‑lived products

How to use this radar

Treat it as a decision aid. For each item we provide: rationale, risks, migration paths, and reference implementations. We update positions based on real project outcomes, so your stack reflects what works, not just what’s trending.

Get in Touch

Whether you need an architectural second opinion, a turnkey delivery team, a safety/compliance uplift, or a critical hire, cpp.ai brings the tooling, process, and talent to deliver with confidence.