A day at IBM in New York

Tue, 31 Mar 2026 18:00:00 -0400

IBM HQ, NYC — March 31, 2026

A small note before I begin

This is a personal reflection, not a recap. I’m deliberately not writing about the sessions, the content, or the specifics of what was discussed — those belong to their owners, not to me. Think of this as a journal entry from a day that meant something, not a newsletter about the day’s substance.

How I got here

About two years ago I was a fresh graduate from Model Engineering College in Kerala, writing far too many late-night side projects, organizing hackathons with my friends, and giving FOSS talks in college auditoriums. The company I work at now, Athena Intelligence, didn’t really exist for me yet. The skyline of New York existed only in YouTube vlogs.

On March 31st, I spent the day at IBM in New York representing Athena. It’s hard to express how far that sentence would’ve been from plausible, even two years ago.

The building is at One Madison Avenue — the Flatiron district, right across from the old Met Life Clocktower. The lobby is all polished stone and sharp angles; the elevators are so quiet they feel architectural. I went up the escalator to the 10th floor with a visitor badge clipped to my jacket, took the turn they told me to take, and suddenly I was in a room with twenty-something enterprise leaders.

I was genuinely nervous.

The Moleskine moment

I arrived early. Found my name on the table. Sat down. Briefly stared.

Aldrin Jenson · Athena Intelligence.

I wasn’t prepared to be affected by something as small as seeing my name on a printed placeholder. But I was. For a few seconds it felt strangely weighty — this is the job, this is the life, you are the person holding this seat today, whether or not the voice in your head agrees.

I didn’t say anything. I just breathed and opened the Moleskine they’d left at each seat.

I took more notes that day than I did in all of my final year of college combined. Not necessarily because the content was transformative — I can’t talk about that anyway — but because the writing itself was a way to stay present. Every time the imposter voice showed up, I just wrote another line. The Moleskine was armour.

What I was thinking about

In quiet moments between conversations, I kept coming back to the same question I’ve been sitting with at Athena for a while now: what do enterprises actually want from AI right now?

My honest answer, from what I see every day in our customer conversations:

They don’t want more tooling. They want fewer surfaces that do more. The CTO who gave me the clearest articulation of this at Athena said: “I have ten AI vendors. I don’t need an eleventh. I need one of them to actually work.”
They don’t want to be wowed. They want to be de-risked. The promise of generative AI for three years has been “look at this amazing demo.” The promise enterprises actually care about is “this will not break, and when it does, you can see why.”
They don’t want autonomy. They want accountability. Autonomy without accountability is a liability. The interesting design problem is giving agents real latitude while keeping the audit trail a human can actually reason about.

None of this is radical. It’s just the stuff I see every week. It was interesting to sit in a room where versions of the same questions were being asked at much bigger scale than Athena’s customer base.

Talking about Athena

I didn’t do a live demo, but I got to talk about what we’re building in several smaller conversations. The piece I came back to most often was our spreadsheet agent — the part of Athena that writes real formulas, respects existing formatting, and produces multiple document artifacts from a single input.

The shape that seemed to land in every conversation:

Describe what you need in one place. Let the agent produce the five artifacts downstream.

That framing is the whole bet, in a sentence. Instead of humans producing a spreadsheet → humans producing a slide → humans producing a memo → humans reconciling all three, the humans describe the intent once and the agent emits the consistent set of downstream outputs. It sounds small. It isn’t. Ask any CFO team how much time they spend reconciling four versions of the same number.

I also mentioned our multi-agent analyst system — coordinated LLM-based agents working in parallel on analytical workflows around the clock. That’s the product I’ve spent most of the past year on. It’s public, it’s been demoed on Twitter, and it’s been publicly endorsed by AssistantUI — but saying it to someone face-to-face in a room like that one has a different quality than posting about it.

I had the quiet thought: the product you made. The product other companies are paying for. You are the one representing it today. That’s a small sentence that takes years to earn the right to think about yourself.

Going solo

The most unexpected thing wasn’t the content — it was the feeling. Walking in alone, as the only Athena person in the room. Realizing for one full day that “the company” in the room was me.

A few times, someone asked me a question I couldn’t answer. I said:

“I don’t know — let me find out and get back to you.”

Which, it turns out, is what everyone in a room like that also says, just with more gray hair and more practice saying it without flinching.

I think the biggest internal shift for me during the day wasn’t about IBM, or Athena, or the market. It was noticing that the difference between me in that room and the more senior people around me was almost entirely composure — not capability. Capability they had, but so did I (for the things in my lane). What they had more of was the willingness to not pretend. To say “I don’t know” without panicking. To ask a dumb-sounding question that turned out to be the exact right one.

That’s the real senior move, I think. Not knowing more. Being okay with not knowing.

Small quiet moment of "okay, this is actually happening."

Meeting Chad

With Chad Jennings — warm, generous, and very good at making a first-timer feel less like a first-timer.

One of the quiet highlights of the day was finally meeting Chad Jennings in person. Chad was generous with his time, thoughtful in his questions, and made it easy for me to stop being nervous. I don’t remember much of what I said in those few minutes, but I remember the feeling of being listened to carefully by someone who didn’t need to. That’s a small thing and also a large thing.

The walk home

We ended the day with a team dinner at the Clocktower — a beautiful restaurant inside the old Met Life tower, right across from the IBM building. I watched the sun set over Madison Square Park, listened to conversations about markets and models and mortgages in no particular order, and quietly realized I wasn’t as nervous as I’d been in the morning.

The walk back to my apartment afterwards was slow. NYC in early April is cold enough to notice but mild enough that you don’t mind walking. I had the kind of quiet that comes after something you were dreading turned out to be good.

Afterthoughts

A few things I was sitting with:

Showing up is most of it. I nearly let imposter anxiety talk me out of being fully present. I’m glad I didn’t. You learn more in one day of being in a room than in six months of reading about what happens in the room.
My job is a privilege. Not every early-career engineer gets to represent a company at a place like this. I want to remember that instead of only the stress of it. There’s a version of me from two years ago who would’ve given anything to sit in that chair, and I owe him the gratitude.
Capability is the floor; composure is the ceiling. Technical skill gets you in the room. What gets you useful in the room is the willingness to be honest about what you don’t know and keep listening anyway.
Gratitude to IBM for the generous welcome — and to Brendon and the rest of the Athena team for the trust. It matters more than I’d say out loud.

That’s the post. No hot takes, no listicle of insights. Just a good day in New York I wanted to remember.

athena on Aldrin Jenson