Hiranya Jayathilaka
hiranya911.bsky.social
Hiranya Jayathilaka
@hiranya911.bsky.social
Building next gen messaging tech at shortwave.com. Formerly at Google. PhD in CS at UCSB. Interested in all things tech.
Santana Row if you have time for a dinner outing and/or shopping. Castro Street area in Mtn View is also really nice. Loads of hiking trails around -- see bahiker.com
April 22, 2025 at 3:01 AM
See you there ☺️
April 8, 2025 at 3:04 AM
Love DS9. Dominion war is one of the best long running storylines of Star Trek canon IMO.
January 3, 2025 at 6:56 PM
This last week I rolled out 3 different experiments for internal testing. By the end of the week I merged them into a single large experiment, and promoted it for public/external testing. If all goes well it will graduate into a bonafide feature next week 🤞
November 23, 2024 at 8:25 PM
Third: A culture of trust and accountability in the team.

Empower and encourage team members to ship early and often. Trust them to do the right thing. And if something goes wrong (we all make mistakes), trust them to take ownership and do the needful to fix things asap.
November 19, 2024 at 5:30 PM
Second: Local testing and dogfooding as a way of life.

Make the entire product stack locally testable end-to-end. Test new features and other critical changes internally by rolling them out just to your team members. Iterate aggressively.
November 19, 2024 at 5:27 PM
I think it's a combination of few things.

First: automating the mundane. Heavily invest into CI/CD. Adopt tools for automated testing and code health enforcement. Build simple flows for one-click roll outs and stress-free roll back.
November 19, 2024 at 5:26 PM
And on the UI side it is quite a thrill to start with some basic @figma.com mocks, and evolve it towards working @react.dev code.
November 19, 2024 at 12:50 AM
Note to self: Write a new article detailing our latest AI infrastructure.
November 13, 2024 at 6:56 PM
In the year since we published that article, we've made major changes to both our app and the AI infrastructure. In fact, we don't use anything described in that article anymore. But I think it's still a great place somebody to start learning about applying #RAG over large, evolving datasets.
November 13, 2024 at 6:55 PM