Victor on Software

@victoronsoftware.com

Software engineering, security, and related topics. Currently building security telemetry and MDM at Fleet Device Management.

Two AI-assisted PRs can look the same: clean diff, passing tests, green CI. But one came from inquiry. The engineer built the mental model. The other came from delegation. The agent made the problem disappear. victoronsoftware.com/posts/compre... #AI #SoftwareEngineering

Comprehension debt: the hidden cost of AI-generated code

AI coding agents can ship working code before the team has built the mental model to maintain it. That gap is comprehension debt.

victoronsoftware.com

AI coding agents can ship working code faster than teams can build the mental model to maintain it. That gap is comprehension debt. The risk isn’t AI writing code. The risk is merging code no human can explain. Article: victoronsoftware.com/posts/compre... #AI #SoftwareEngineering

Comprehension debt: the hidden cost of AI-generated code

AI coding agents can ship working code before the team has built the mental model to maintain it. That gap is comprehension debt.

victoronsoftware.com

Tried to build a modular monolith. Three compromises later, the "modular" part disappeared. 🫠 Turns out team buy-in isn't enough. You need governance and an architect who doesn't walk away. First attempt. Lessons learned. www.youtube.com/watch?v=lrM3... #SoftwareArchitecture #TechnicalDebt

The modular monolith that wasn't

Our first attempt at modularizing a 500K-line codebase taught us more about team dynamics than architecture Full article: https://victoronsoftware.com/posts/modular-monolith-attempt/

youtube.com

🤖 Can AI agents really make multitasking more productive, or is it just a trap? We broke it down in this video: when to single-task, when to multitask, and how to keep your agents on track. 📺 Watch here 👉 youtu.be/HSDrhYvE3g0 #AI #AgenticAI #DeveloperExperience

Multitasking with AI agents: When it works and when it fails

A guide to balancing focus and efficiency when working with AI agents, including tips on single-tasking vs multitasking Full article:…

youtu.be

We used to think OpenTelemetry was just for prod. Turns out, it’s a dev tool too. 🛠️ Why developers should use OpenTelemetry in dev: → victoronsoftware.com/posts/opente... See how we: ✅ Trace API calls ✅ Debug SQL ✅ Correlate logs ✅ Catch errors early #OpenTelemetry #DevTools #Observability

Why developers should use OpenTelemetry in dev

Learn how using OpenTelemetry during development improves observability and prepares your app for production success

victoronsoftware.com

Will AI agents replace software developers? 🤖 Not anytime soon. They’re great at 💻 implementation & 🧪 testing but not in other areas. Even full code automation = only 43% productivity gain. Full breakdown 👉 victoronsoftware.com/posts/will-a... #AI #AIAgents #DevTools #SoftwareEngineering #LLMs

Will AI agents replace software developers?

A realistic look at how AI coding agents are reshaping software development and what engineers need to do to stay ahead

victoronsoftware.com

🔐 mTLS vs HTTP Message Signatures: which should you use? We break down the tradeoffs for device enrollment & secure APIs. ✅ How they work ⚖️ Pros & cons 📊 Comparison table 🆕 Why RFC 9421 matters 👉 victoronsoftware.com/posts/mtls-v... #CyberSecurity #mTLS #ZeroTrust #SysAdmin #EndpointSecurity

mTLS vs HTTP signature faceoff: securing your APIs

How mTLS and HTTP signatures work, where they fit best, and how to choose the right one for your architecture

victoronsoftware.com

Track engineering metrics with real power—for free 💥 ✅ GitHub Actions ✅ BigQuery ✅ Grafana We built a dashboard with moving averages, filters, & drill-down links—all on free-tier tools. 👉 victoronsoftware.com/posts/engine... #EngineeringManagement #Metrics #SoftwareDevelopment

Full-featured engineering metrics—for free

Use GitHub, Grafana, SQL, and free-tier infrastructure to power your engineering benchmarks

victoronsoftware.com