Blog

Notes on systems, safety, and AI.

Write-ups from building reliable systems and practical AI tools. Sample content — only the founder note is a full post so far.

Founder · Aug 2026 · 6 min read

Why I started PerunTeCH

Two companies, one motivation: building systems my son's generation can rely on. The honest story of going from safety-critical SoCs to founding an AI tools company.

Read the post →

AI & LLMs · Jul 2026 · 8 min read

Audit trails for AI-written work

If an LLM drafted it, you should be able to prove what it knew and when. How we bring traceability discipline from ISO 26262 into AI-assisted workflows.

Teaser only · full post coming

Safety · Jun 2026 · 7 min read

Safety standards as design inputs, not paperwork

Compliance follows naturally from architecture when you treat the standard as a requirement source. Lessons from automotive SoCs and rail.

Teaser only · full post coming

AI & LLMs · May 2026 · 5 min read

Checkpoints beat magic: designing assistant flows

Step-by-step flows with clear confirmations outperform end-to-end automation for high-stakes tasks. What our prototypes taught us.

Teaser only · full post coming

Method · May 2026 · 9 min read

The Verified-Record Method

Representing myself in court taught me a discipline: write everything down, with dates and times. It now ships inside our AI Engineering product.

Teaser only · full post coming

Systems · Apr 2026 · 6 min read

Keeping embedded systems debuggable

Connecting low-level firmware to higher-level tooling without losing debuggability — patterns from MSP430 hobby work to production SoCs.

Teaser only · full post coming