01 — BeforeWhat I Expected
I'd been to online events. Watched conference talks on YouTube at 1.5x speed, skimming for the useful parts and closing the tab the moment slides got dense. A live conference felt like the same thing but with worse seating and no pause button. I went primarily because a senior at college had a spare ticket and made it sound important. I had no agenda, no plan, no list of speakers I wanted to see.
My expectation was simple: sit in a room, absorb information, come back smarter. I was almost entirely wrong about what that would mean.
A mid-sized tech conference — a few hundred attendees, two tracks running in parallel, a mix of industry engineers, researchers, and students. Talks on ML infrastructure, developer tooling, and system design. The kind of event that looks comprehensive on paper and overwhelming in person.
02 — Day OneThe Talks Were Not the Point
The first session I attended was on distributed systems at scale. The speaker was clearly expert — calm delivery, precise language, slides that didn't try to be clever. I took notes. I understood maybe 60% of it in the room and another 20% on the train home after looking things up. The remaining 20% is still somewhere in a notebook I haven't opened since.
But the talk itself wasn't what stayed with me. What stayed was the conversation immediately after, standing at the coffee station with two people I didn't know — a backend engineer from Bangalore and a final-year student from Hyderabad. We were all confused by the same part of the talk. That shared confusion was more valuable than the session itself. We spent fifteen minutes unpacking it together, and I understood more from that exchange than from the forty minutes preceding it.
The talks give you the vocabulary. The conversations give you the understanding. You need both — but nobody tells you the second part is where most of the learning actually happens.
I started paying attention to this pattern. Every good learning moment that day had the same structure: a talk surfaces a concept, a hallway conversation forces you to articulate your confusion, and the act of articulating it is where it clicks.
03 — The GapsWhat Happens Between Sessions
I had assumed the schedule was the product. The ten-minute breaks were logistics, not content. I was wrong about this in a way that felt obvious only in retrospect.
Got talking to a developer who had shipped a feature I'd read about in a changelog but never thought deeply about. He explained the decision-making behind it in two minutes — the constraints, the tradeoffs, the thing they almost did differently. No talk would have given me that level of specificity.
Sat in the wrong room for five minutes before realising I was at the wrong talk. Stayed anyway. The speaker was covering something I'd had no interest in — developer experience tooling — and by the end I had three new ideas for improving my own workflow. The accidental session was the best one I attended.
Spent an hour avoiding the networking session because I didn't know what to say to strangers. Went in for ten minutes because I felt guilty. Had one genuinely interesting conversation about how small teams make architecture decisions without formal process. Left wishing I'd gone in earlier.
04 — The ReframeWhat I Got Wrong About Networking
I'd built networking into something transactional and uncomfortable — an activity where you exchange business cards and try to make yourself sound impressive. That framing made me avoid it for most of day one.
The reframe that worked: a conference is a temporary environment where asking genuine questions is socially expected and encouraged. Everyone is there because they're interested in something. "What brought you here?" is not small talk — it's a direct route to a conversation about what actually matters to someone. I had better conversations once I stopped trying to network and started trying to understand what the person in front of me was working on.
Stop trying to be interesting. Start trying to be interested. The conversations that followed this shift were longer, more specific, and left me with things I actually thought about afterward.
05 — The TalksWhat Makes a Good One
Watching talks in person gave me a clearer sense of what separates a good one from a forgettable one — and it's almost never the technical depth.
The talks I retained were the ones where the speaker was honest about what they didn't know, where they got it wrong, or where the thing they built didn't work the way they expected. Uncertainty handled well is more instructive than expertise performed confidently. One speaker paused mid-slide, said "I'm actually not sure this is the right approach anymore," and spent three minutes thinking through it in real time. That moment was more educational than the rest of the session.
06 — TakeawaysWhat I Came Back With
Every moment of confusion in a talk is a question worth asking out loud afterward. The people who asked questions consistently had better follow-up conversations.
The scheduled content is the frame. The unscheduled conversations are where ideas actually form. Optimise for time between sessions, not just coverage of sessions.
At least once, walk into a session outside your area without an agenda. Accidental exposure is frequently more valuable than another talk on something you already care about.
Dense notes filled two talks I remembered almost nothing from. By day two I was writing one sentence per session — the thing that actually changed how I thought about something. Those are still useful.
The people I'd read about were, in person, mostly curious and uncertain about different things. Expertise is a practiced comfort with not knowing — and knowing what to do next anyway.
I went expecting to download information. I came back with something less tangible and more useful — a clearer sense of how people who are good at this actually operate, what they're uncertain about, and how they talk to each other when nobody is performing. That's not something you get from a YouTube playlist. It took being in the room.