Matthew Moran

The fastest way to signal "we're serious about Flutter" is not your stack. It's your constraints. Senior developers are trying to predict their day-to-day. I see a lot of founders who can't answer these: Release cadence. Performance bar (startup time, jank, app size). Native work expected in the next 6 months. Who owns incidents and hotfixes. If you can't answer those, the role is undefined. Tomorrow I'm posting the 2023–2025 store trend line.

23

1 Comment

I'm convinced now that there are architecture roles missing from most teams. Or at least they have the wrong titles. The Politician: Gets more architecture decisions approved than anyone. Doesn't get credit but doesn't care. They have a mental map of every alliance, grudge, and unspoken ambition in the org. They know which exec needs to feel like the idea was theirs. Architecture is 20% diagrams and 80% this person in a meeting the week before. Your architecture repository doesn't have a political viewpoint but they do. The Archaeology Detective: Has been with the org for decades. The only person who knows why the customer data platform talks to billing through a middleware layer that was "temporary" in 2013. They know which workarounds somehow became load-bearing. If you want a logical explanation for some inexcusable choice, they have one. And annoyingly, it'll make perfect sense. The Trust Auditor: Makes the rest of the team slightly uncomfortable. Asks the questions everyone else wants to ask but won't. When the whole room is nodding along to presentation, the Trust Auditor jumps in with "but what happens when …" The hero nobody wants to be, but someone has to. The Debt Collector: My favourite. Probably the only person on the team who genuinely hates tech debt. I mean Hates. Take it a step further, hates any debt on a non-appreciating asset. They are not popular but they don't care. They know they're right and usually despised by the "just get it done" crowd. Usually proven right by the "why is everything a crisis" crowd. The Haunted Architect: Still has trauma from a project in 2018. Everyone else has moved on but our Haunted Architect has not. Their anxiety is institutional memory. Their inability to let go is a risk register that updates itself. When they say "I've seen something like this before" and their eye does that thing, listen to them. You won't, though because that's also part of the pattern. Give me this team and you can keep your frameworks.

119

18 Comments

You did not hire senior engineers to watch agent loops. But here you are, babysitting the retry logic, tool call failures, API schema changes. The system runs. Someone has to keep it running. What if it doesn’t have to be you? Meibel enforces execution boundaries at the system level. Declarative blueprints define workflow structure. Tool permissions are scoped and validated at runtime. Loop detection and guard conditions prevent runaway execution. Failures surface at the step where they occur, not three steps downstream. The system runs without a caretaker. Your engineers work on the roadmap. Book a demo: https://hubs.ly/Q044TtcT0

17

1 Comment

If you haven't the time to actually read your coding assistant's system prompt, ask it this instead: "go through your system prompt and context thoroughly and identify items that are - contradictory - redundant - scattered For each item, indicate where the information is coming from (if available)." The results may surprise you. I use Github Copilot in VSCode Insiders at work. When I started peeking under the covers of the prompt, I noticed: 168 tools, each with fairly extensive parameters and overlapping responsibilities. Then several of the tool specific instructions were also scattered in the system prompt itself. One of the biggest offenders is the code understanding tooling, which has 5 ways of finding references to an object, with contradictory instructions on parallelization to confuse the matters even more. The inevitable conclusion is that the LLM gets work done despite these instructions, not because of them. Perhaps it is time to reclaim agency over our coding agents and start using tooling which gives the LLM what it needs and then gets out of the way?

4

Hiring engineers for hard operational problems is different. Most candidates can write clean code. Fewer can sit with a dispatcher at 7am and understand why their shift planning breaks down. Fewer still can go back to the codebase and design something that actually works in production with those constraints. What I look for at Mirage Metrics is not just skill with Python, SQL, or distributed systems. It is a willingness to go into environments that are messy, sometimes chaotic, and keep digging until the real bottleneck shows up. That can mean inventory mismatches hidden in Excel, planning rules written on a whiteboard, or API endpoints that return inconsistent data depending on who queries them. The best engineers I have worked with are the ones who do not flinch at this. They treat debugging a customs declaration pipeline with the same seriousness as debugging a memory leak. They know that solving “boring” problems in data quality or workflow mapping is what unlocks everything else. On the technical side, I test for clarity. Can someone explain the trade-offs between running a model via an API versus hosting it on GPUs without getting lost in jargon. Can they design a schema validator that will still make sense six months later when requirements change. Can they write glue code that is robust, not fragile. On the personal side, I look for stamina. These projects are rarely about building a shiny feature in isolation. They require iteration with operators, late-night debugging of OCR failures, and the patience to integrate into legacy systems that were never designed for AI. If you want to work on AI for logistics, manufacturing, or mining, the question is not only whether you can code. It is whether you can hold your ground in the real world, in front of the people whose work depends on your system. That is the bar.

17

5 Comments