Skip to main content

tech radar · v2 · august 2026

What I'm betting on, and what I'm not

Years ago I launched my first company's Tech Radar to give engineering teams a shared formulary for what to adopt, trial, assess, and hold. This is the personal version. 86 blog posts and one year of writing, distilled into the bets I am willing to defend out loud.

last reviewed August 29, 2026 · click any blip or any item below for the argument behind it

ADOPTTRIALASSESSHOLDEngineering PracticesAI Tools & ArchitectureLeadership PosturesIndustry Narratives12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182
Adopt — use today. default in new workTrial — in production. convinced of directionAssess — watching. not yet bet onHold — refuse, or warn against

quadrant 01

Engineering Practices

How software actually gets built, reviewed, and shipped in an agentic SDLC.

Adopt

Use today. Default in new work.

  • 01Spec-driven development
  • 02Standards in files, not heads
  • 03Architecture as agent guardrails
  • 04Pipeline-first delivery
  • 05Verification-first development
  • 06Fail to unknown, never to green
  • 07Conformance suite as executable oracle
  • 08Name an owner for every agent
  • 09Separation of duties for agents
  • 10Context Spine / Intent Stack
  • 11Enforced design systems, not voluntary ones

Trial

In production. Convinced of direction.

  • 12Answer-key-first modernization
  • 13Operability in the spec
  • 14Spec review as the new code review
  • 15Multi-agent specialist workflows
  • 16Prompt review as quality gate
  • 17Workflow inversion over agent loops

Assess

Watching. Not yet bet on.

  • 18Dark factory model for software
  • 19Cross-team multi-agent workflow sharing

Hold

Refuse, or warn against.

  • 20Vibe coding without specs
  • 21Kubernetes for one service
  • 22Agent teams as a default
  • 23Line-by-line review as primary quality bar

quadrant 02

AI Tools & Architecture

The seam between local and frontier, the patterns that compound, the ones that do not.

Adopt

Use today. Default in new work.

  • 24Claude Code as build system
  • 25Hybrid local + frontier inference
  • 26MCP as integration substrate
  • 27AI providers like cloud providers
  • 28Design for sovereign model risk
  • 29Own the learning loop, not the model

Trial

In production. Convinced of direction.

  • 30Open-weight models for regulated workloads
  • 31Persistent personal AI agents (Roci)
  • 32Custom-trained frontier models (Nova Forge style)
  • 33Design the platform for the agent persona
  • 34Context as a platform capability
  • 35Invisible AI over the portal chatbot
  • 36LLM-consumable design systems

Assess

Watching. Not yet bet on.

  • 37Intent interface over the golden path
  • 38Fully autonomous coding agents
  • 39MCP gateway / RBAC patterns
  • 40AI design tools beyond Figma

Hold

Refuse, or warn against.

  • 41Multi-cloud as a hedge
  • 42Prompt-whispering as a discipline

quadrant 03

Leadership Postures

How engineering leaders should show up, what they should reward, and what they should refuse.

Adopt

Use today. Default in new work.

  • 43Stay close to the craft
  • 44Principles over personalities
  • 45Accountability flows upward
  • 46Clarity beats consensus
  • 47Turn complexity into leverage
  • 48Servant leadership in tech
  • 49Invest in framers, not finishers
  • 50Judgment as the scarce resource
  • 51Bargain honestly with disruption
  • 52Value per verified outcome

Trial

In production. Convinced of direction.

  • 53FinOps as a core engineering competency
  • 54Build for succession
  • 55Designed apprenticeship over accidental
  • 56The AI Operating Ledger
  • 57Spend the saved hour on purpose

Assess

Watching. Not yet bet on.

  • 58Pairing on specs and reviews

Hold

Refuse, or warn against.

  • 59Mandate-driven AI adoption
  • 60Building monuments to yourself
  • 61Headcount-first AI reorgs
  • 62Personal agents as AI strategy
  • 63PowerPoint-only leadership

quadrant 04

Industry Narratives

The stories the industry tells itself. Some are true. Most are someone's quarterly earnings call dressed up in a hoodie.

Adopt

Use today. Default in new work.

  • 64"Specs beat vibes"
  • 65"The bottleneck is never the stack"
  • 66"Agents advise; humans decide"
  • 67"From toil to judgment"
  • 68"Trust the gate, not the actor"

Trial

In production. Convinced of direction.

  • 69"Trust compounds faster than velocity"
  • 70The builder’s bargain
  • 71"The frame is the bottleneck"
  • 72Open weights as commoditization strategy

Assess

Watching. Not yet bet on.

  • 73AI infrastructure capex bubble
  • 74Open-source funding under AI
  • 75Capability as a governed resource
  • 76The rewrite taboo was a price tag
  • 77Intelligence as infrastructure

Hold

Refuse, or warn against.

  • 78"Workforce optimization" language
  • 79"AI is coming for your job"
  • 80"AI replaces engineers"
  • 81The salary-for-subscription swap
  • 82Vendor-as-operating-model