Hmm, the world is a different place now. I used to write a lot of code to build features, but now I’m doing something else, something that’s more technical in nature and design oriented. I unblock people, resolve production issues, take on customer requests, and focus on the system’s durability. But I still like building shiny new things. It makes me feel like I’m doing something, and that something is somehow generating revenue.
So yeah, what did I do in October? I took a workshop on “AI/Prompting for Non-Tech People.” I added a lot of memes to my slides, and I think it went well. By “non-tech”, I mean people who don’t write code for a living. I think it was the highlight of the month. Why? Because it gave me visibility in the org.
The rest of the month went into fixing things - deployments, CI/CD pipelines, retry mechanisms, writing documents for POCs, and a few other tasks.
I also got two courses that I need to complete before Feb 2026. It’s a company requirement. The courses are about ArgoCD and building AI agents. What are AI Agents? Why are AI Agents? Will the AI Agents charge money to get things done? Agents. I’ve always thought I’m more interested in base system engineering like distributed systems, kernels, compilers, and all that. I always wanted to work on long-term problems, and now that I finally got one, well, it’s long. There’s no clear solution, just a bunch of trade-offs I’m still thinking about. It’s about decoupling a join dependency across different databases. Sounds pretty simple, right? But decoupling systems also adds multiple failure points, so you have to add restart mechanisms to make sure events don’t get missed. And even if they do, the system should be able to correct itself. Self-healing systems. OMG.
Writing and committing code feels like the best path to job satisfaction, but how much code can you really write? To move ahead in your career as a techie, you have to eventually step back from writing code and focus on bigger problems. I see many people in Staff Engineer roles, and I don’t think they’re writing much code every day. Maybe reviewing a lot of it, but not necessarily writing it.
Anyway, I also fixed a production issue. The issue was pretty silly, instead of running a create-or-update query, only the update query was being executed, breaking the webhook ingestion flow. Lack of testing leads to such problems. As a dev, one needs to make testing easier. There were no unit tests, integration tests, or anything. The gap was wide enough for bugs to slip through. And when you depend on third-party services, you’ve got to cover all your tracks. Things can break in weird ways anytime.
Oh, and I also optimized a legacy API. It was running about 170 queries and taking 15 seconds to complete. There were no tests or assertions of any kind, so I created some, then optimized it down to 13 queries. Now it takes just one or two seconds. Ehehe.
I also audited production data and created reports explaining what the gaps were, why they existed, and how they could be fixed. I spent a fair bit of time on calls too. As a dev, I don’t enjoy long meetings, but brainstorming can be fun sometimes. I also pointed out a few cases where our servers’ disks could fill up because of some video processing tasks. And I moved the report creation and download feature I built to S3. Once the reports are created, they’re uploaded there and that’s it :P Not much, but it was honest work.
I worked on bugs and a bunch of other things. I think my scope as an engineer is changing. I don’t know if that’s good or bad. I am becoming a better engineer though. Still, I want to create an algorithm that makes distributed systems more robust. Or build a kernel. Or make some OSS contribution to Linux. All of this feels like a pipe dream right now, but maybe someday.
I also need to update my website. I don’t know what to write there. I want to make it look nice, maybe give it a retro vibe. About OSS contributions - I already look at code from Monday to Friday, so looking at more code on weekends feels like a drag. I love coding, but sometimes you need to clear your mind. A lot of solutions come to you when you sleep and wake up with a ground-breaking idea for your problems.
This blog is slowly turning into a feelings blog. Ugh. I also want to try sales, just for fun, to see how it all works. I mean, we already sell ourselves in interviews, but selling an enterprise B2B SaaS must be a whole different experience.
I’ve also lost all my interviewing skills, and I was never really good at DSA-style interviews anyway. Man, sometimes the future looks bleak. Damn, this was supposed to be a technical post, but now we’re walking the thin line between real and delusional.
