tech radar · v2 · june 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. Sixty-three blog posts and one year of writing, distilled into the bets I am willing to defend out loud.
last reviewed June 21, 2026 · click any blip or any item below for the argument behind it
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
- 06Name an owner for every agent
- 07Context Spine / Intent Stack
- 08Enforced design systems, not voluntary ones
Trial
In production. Convinced of direction.
- 09Spec review as the new code review
- 10Multi-agent specialist workflows
- 11Prompt review as quality gate
- 12Workflow inversion over agent loops
Assess
Watching. Not yet bet on.
- 13Dark factory model for software
- 14Cross-team multi-agent workflow sharing
Hold
Refuse, or warn against.
- 15Vibe coding without specs
- 16Kubernetes for one service
- 17Agent teams as a default
- 18Line-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.
- 19Claude Code as build system
- 20Hybrid local + frontier inference
- 21MCP as integration substrate
- 22AI providers like cloud providers
- 23Design for sovereign model risk
Trial
In production. Convinced of direction.
- 24Open-weight models for regulated workloads
- 25Persistent personal AI agents (Roci)
- 26Custom-trained frontier models (Nova Forge style)
- 27LLM-consumable design systems
Assess
Watching. Not yet bet on.
- 28Fully autonomous coding agents
- 29MCP gateway / RBAC patterns
- 30AI design tools beyond Figma
Hold
Refuse, or warn against.
- 31Multi-cloud as a hedge
- 32Prompt-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.
- 33Stay close to the craft
- 34Principles over personalities
- 35Accountability flows upward
- 36Clarity beats consensus
- 37Turn complexity into leverage
- 38Invest in framers, not finishers
- 39Bargain honestly with disruption
- 40Value per verified outcome
Trial
In production. Convinced of direction.
- 41FinOps as a core engineering competency
- 42Build for succession
- 43Servant leadership in tech
- 44The AI Operating Ledger
Assess
Watching. Not yet bet on.
- 45The apprenticeship problem
- 46Pairing on specs and reviews
Hold
Refuse, or warn against.
- 47Mandate-driven AI adoption
- 48Building monuments to yourself
- 49Headcount-first AI reorgs
- 50Personal agents as AI strategy
- 51PowerPoint-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.
- 52"Specs beat vibes"
- 53"The bottleneck is never the stack"
- 54"Agents advise; humans decide"
- 55"From toil to judgment"
- 56"Trust the gate, not the actor"
Trial
In production. Convinced of direction.
- 57"Trust compounds faster than velocity"
- 58The builder’s bargain
- 59"The frame is the bottleneck"
Assess
Watching. Not yet bet on.
- 60AI infrastructure capex bubble
- 61Open-source funding under AI
- 62Capability as a governed resource
Hold
Refuse, or warn against.
- 63"Workforce optimization" language
- 64"AI is coming for your job"
- 65"AI replaces engineers"
- 66The salary-for-subscription swap
- 67Vendor-as-operating-model