tech radar · v2 · july 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-six blog posts and one year of writing, distilled into the bets I am willing to defend out loud.
last reviewed July 2, 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
- 06Fail to unknown, never to green
- 07Name an owner for every agent
- 08Context Spine / Intent Stack
- 09Enforced design systems, not voluntary ones
Trial
In production. Convinced of direction.
- 10Spec review as the new code review
- 11Multi-agent specialist workflows
- 12Prompt review as quality gate
- 13Workflow inversion over agent loops
Assess
Watching. Not yet bet on.
- 14Dark factory model for software
- 15Cross-team multi-agent workflow sharing
Hold
Refuse, or warn against.
- 16Vibe coding without specs
- 17Kubernetes for one service
- 18Agent teams as a default
- 19Line-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.
- 20Claude Code as build system
- 21Hybrid local + frontier inference
- 22MCP as integration substrate
- 23AI providers like cloud providers
- 24Design for sovereign model risk
Trial
In production. Convinced of direction.
- 25Open-weight models for regulated workloads
- 26Persistent personal AI agents (Roci)
- 27Custom-trained frontier models (Nova Forge style)
- 28LLM-consumable design systems
Assess
Watching. Not yet bet on.
- 29Fully autonomous coding agents
- 30MCP gateway / RBAC patterns
- 31AI design tools beyond Figma
Hold
Refuse, or warn against.
- 32Multi-cloud as a hedge
- 33Prompt-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.
- 34Stay close to the craft
- 35Principles over personalities
- 36Accountability flows upward
- 37Clarity beats consensus
- 38Turn complexity into leverage
- 39Servant leadership in tech
- 40Invest in framers, not finishers
- 41Bargain honestly with disruption
- 42Value per verified outcome
Trial
In production. Convinced of direction.
- 43FinOps as a core engineering competency
- 44Build for succession
- 45The AI Operating Ledger
Assess
Watching. Not yet bet on.
- 46The apprenticeship problem
- 47Pairing on specs and reviews
Hold
Refuse, or warn against.
- 48Mandate-driven AI adoption
- 49Building monuments to yourself
- 50Headcount-first AI reorgs
- 51Personal agents as AI strategy
- 52PowerPoint-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.
- 53"Specs beat vibes"
- 54"The bottleneck is never the stack"
- 55"Agents advise; humans decide"
- 56"From toil to judgment"
- 57"Trust the gate, not the actor"
Trial
In production. Convinced of direction.
- 58"Trust compounds faster than velocity"
- 59The builder’s bargain
- 60"The frame is the bottleneck"
Assess
Watching. Not yet bet on.
- 61AI infrastructure capex bubble
- 62Open-source funding under AI
- 63Capability as a governed resource
Hold
Refuse, or warn against.
- 64"Workforce optimization" language
- 65"AI is coming for your job"
- 66"AI replaces engineers"
- 67The salary-for-subscription swap
- 68Vendor-as-operating-model