Andy
@viewerisland
Followers
351
Following
292
Media
16
Statuses
332
Building Osmosis (YC W25)
San Francisco, CA
Joined June 2014
Feeling like I have a super power when I can read CSDN blogs and random alibaba developer docs. Info density is unmatched when compared to US articles + deep research tools
0
0
1
Had a quick thought thinking about a new type of reward function for a RL experiment on the commute home, so I launched one on my phone!
2
0
11
Recent years have seen GPUs become the dominant chip within a system, overtaking CPU importance. GeminiFS is a natural progression of this change.
1
0
3
It was a really good session! Amazed by the turn out and shared some really valuable notes, check out the recordings if you missed the event!
We brought together 200+ founders, researchers, and engineers on Saturday for RL IRL at @ycombinator HQ! At @Osmosis_AI, our belief is that RL is the missing piece for AI agents - so we had some friends share how they're applying RL today! 🧵of session recordings below:
0
0
10
Hey that graphic looks like it's got some numbers on it, wonder where it's from? But really, come to the event! I think it'll be a really helpful session on designing RL ready AI apps.
Sharing the full schedule for RL IRL (Saturday at @ycombinator) below! We're finalizing the guest list tonight - this is the last chance to RSVP. (Also, a sneak peek at some of the @Osmosis_AI merch we're handing out!)
0
0
7
Let's define what it means for an application company to be "RL ready"!
Super excited for RL IRL this Saturday at @ycombinator, we're at 500+ registrations! We (@Osmosis_AI) are co-hosting a session with @greptileai on designing RL-ready products - Other sessions in thread 👇
0
0
9
Super excited for RL IRL this Saturday at @ycombinator, we're at 500+ registrations! We (@Osmosis_AI) are co-hosting a session with @greptileai on designing RL-ready products - Other sessions in thread 👇
11
12
108
Thank you so much @rauchg for being an early supporter of @Osmosis_AI - been amazing to explore the opportunities for RL at Vercel!
@_WEEXIAO @Osmosis_AI congrats, been enjoying working with you guys
1
1
43
Thrilled to be a part of the journey @Osmosis_AI! Applications have the data, now it's time to use it. Lfg @_WEEXIAO @viewerisland 🤗
4
6
71
Osmosis is a platform for companies to fine-tune models that outperform foundation models with reinforcement learning. Better, faster, and cheaper. Learn more at https://t.co/2d3bSeYwvV
5
1
37
We've raised $7M to help companies build AI agents that actually learn and work. @Osmosis_AI is a platform for companies to fine-tune models that outperform foundation models with reinforcement learning. Better, faster, and cheaper.
137
91
639
I met my co-founder @_WEEXIAO in middle school over ten years ago. This year, we went through YC, and raised $7M from amazing investors. Today, we officially launch @Osmosis_AI - a reinforcement fine-tuning platform purpose built for continuous improvement. Learn more below!
5
4
34
We're already at 200+ registrations for RL IRL! We'll start going through and accepting folks this weekend - RSVP soon!
Announcing RL IRL: a forum for researchers/engineers focused on applied RL held at YC on 10/25. We're (@Osmosis_AI) excited to be co-hosting with @ycombinator, @shaneguML, @reductoai, @greptileai, and @encord_team! We’re capping at 150 - register below to attend and/or speak!
3
3
22
Announcing RL IRL: a forum for researchers/engineers focused on applied RL held at YC on 10/25. We're (@Osmosis_AI) excited to be co-hosting with @ycombinator, @shaneguML, @reductoai, @greptileai, and @encord_team! We’re capping at 150 - register below to attend and/or speak!
9
10
84
It's been nearly a year since @viewerisland and I started @Osmosis_AI! Yesterday, we celebrated our first real office (+ my birthday). Goodbye apartment office... It was amazing to catch up with old friends and make new friends! Couldn't be more excited for next week 👀
8
7
32
With the new SGLang kernel implementation, you can suddenly have a deterministic RL CICD pipeline. Imagine commenting "at step 10 the reward was 0.003 lower and that's not normal"
0
0
2