Leading GTM @AirbyteHQ | Building data infrastructure for AI agents.

San Francisco, CA
Joined June 2009
Mario Moscatiello retweeted
Cloud → Multi-cloud → Hybrid → Private/Sovereign
90
77
32
644
266,732
Mario Moscatiello retweeted
Today we are releasing Laguna S 2.1. At 118B total parameters, with 8B active per token, it does the work of models several times its size on agentic coding. It is remarkably persistent across long-horizon tasks. And it is small enough to run on a single NVIDIA DGX Spark. It is far more capable than anything we have created before, and I think it redefines what a model in its weight class can do. Laguna S 2.1 is an important model for Poolside. What it represents is even more important. If, five years ago, I had read a book that said that by 2030 everything economically valuable, scientifically interesting, and personally meaningful would be built on intelligence contracted from three or four companies, I would have called it dystopian science fiction. We are at a fork in the road of what kind of world we can have. I believe intelligence should and will become a commodity. The question is whether that intelligence comes from three companies, or from many people who can build it, own it, and shape it. The open ecosystem will not win by being the best in its own category. No one cares who is king of the open-source kingdom. People want the best intelligence for the task they are trying to do, with the right balance of quality, speed, cost, and control. If we want a different future, open models have to be on par with, or better than, their closed equivalents. Laguna S 2.1 is a meaningful step in that direction: capable enough to compete far above its weight class, efficient enough to run on hardware you can own, and open-weight so anyone can build on it. Open-weighting our models is the contribution we can make today toward a world where intelligence can be built and owned by many. And we will keep doing it. I am very proud of this team’s work. A big shout out to everyone at Poolside who made this possible, from infrastructure and data to architecture, pretraining, post-training, evaluations, and inference. Laguna S 2.1 is available today under the OpenMDW-1.1 license, with weights on Hugging Face and access through OpenRouter and our API. We are building toward a future where the most capable intelligence in the world can be owned and shaped by anyone. Laguna S 2.1 is one step. We are going to keep building until that future exists. poolside.ai/blog/introducing…
164
191
107
1,729
338,873
Mario Moscatiello retweeted
Today we're launching Airbyte Agents. We built this product to solve the discovery problem for production-grade agents. Every API assumes you already know what you want to query. Endpoint, object ID, fields, etc. Agents start one step earlier. They need to discover which entities matter before they can fetch or act. That discovery gap is where most agent architectures break. The Context Store closes it. Pre-materialized business data. Up to 80% fewer tokens. 40% fewer tool calls. Deployable via MCP, Python SDK, or our no-code builder (in research preview). 50+ open source agent connectors for the most popular tools, with more coming each week. We’re early, and the product is getting better every week. We can’t wait for you to test it! youtube.com/watch?v=xvzIUNRI…
17
16
4
67
28,276
Mario Moscatiello retweeted
Most companies calling themselves AI-native added an AI feature. That is not the same thing. Talked about this with @chetanp at @StartupGrind yesterday. Being AI-native shows up in three places: how you operate, what you build, and how you make money. Thanks to SG for having us!
2
5
13
2,875
Excited for this!
Today @poolsideai is releasing Laguna M.1 & Laguna XS.2, our latest generation models and first public models We started Poolside because we believed that to build truly capable coding agents, you need to own the full stack: data, training, reinforcement learning, inference. These models are the first result of that work, and we’re making them available to everyone
5
103
Mario Moscatiello retweeted
Today @poolsideai is releasing Laguna M.1 & Laguna XS.2, our latest generation models and first public models We started Poolside because we believed that to build truly capable coding agents, you need to own the full stack: data, training, reinforcement learning, inference. These models are the first result of that work, and we’re making them available to everyone
44
40
5
389
53,174
This 👇
The number one negative thing that smart people do is overvalue optionality. By limiting your options, you allow yourself to be clear and focused. Focus is the act of eliminating options.
1
3
416
Mario Moscatiello retweeted
Open vs. Closed battles are cool and all, but the real hard core innovation begins when the game is Open vs. Open.
3
6
42
4,133
Mario Moscatiello retweeted
🆕 Excited to roll out our Stripe wizard after 3 months of dogfooding + private beta! → Like a SaaS metrics tool, we give you an ARR dashboard in 1 click → Unlike a SaaS metrics tool, it's in Causal so is fully customisable Try it out :) go.causal.app/arr
2
1
43
5,177
Mario Moscatiello retweeted
🎤 Very pleased to announce Causal 2.0: Modern Financial Planning for Startups Causal is powerful but always had a steep learning curve. We spent the last 6 months re-thinking this — you can now get value in ~2 mins, and become a power-user over time. Here's how we did it 👇
19
19
5
175
39,426
Bigger than Blackwell is @nvidia’s "NIM" for enterprises. Nvidia Inference Microservices. Enterprise SaaS & SW Platforms (ie Adobe, SAP) & Data Platforms (ie Cloudera, Cohesity, & SnowBricks) write once across the hybrid multi-cloud Infrastructure (ie AWS, Dell) and Model Builders (ie Meta, Microsoft). Integrated into "NV AI Enterprise 5.0". This is the ultimate embrace & extend play and will be very hard to replicate. This is bigger than CUDA for enterprises. $NVDA
3
49
8
260
41,606
Mario Moscatiello retweeted
Moores law? 30x every 10 yrs. DGX GB200? 3600x the compute of the DGX delivered to Open AI to train GPT1 in 2016. And could train GPT4 w 1/10 the GPUs & 1/4 the power. And still 2 yrs to go. Living in a New Age. $NVDA #blackwell 🚀
21
71
15
572
85,482
Containerization is upon us.
2
99
Mario Moscatiello retweeted
Step 1: Decide you might want to buy it Step 2: Watch every YouTube video that's ever been created about it
954
5,435
1,068
56,591
2,982,940
Mario Moscatiello retweeted
EXCLUSIVE! 🎙️🎙️🎙️ @poolsideai is a very ambitious Generative AI for code startup that’s already raised $126M For their FIRST long-form conversation, CTO @eisokant joined us on the pod for a glimpse as what he and @jasoncwarner have been building 👇LINKS IN NEXT POST👇👇
4
10
1
49
19,166
Mario Moscatiello retweeted
i love @cartainc but starting to see a lot of startups move off of Carta because they charge $$$$. anyone else seeing this?
93
7
4
244
144,157
Mario Moscatiello retweeted
Brian Halligan @bhalligan hypothesized a entire genre of marketing and built a $20B company out of that idea: @HubSpot On this week’s episode of World of DaaS, we discuss: 💙 OpenAI board 🤝 what makes good (and bad) boards 🌪️ ChatGPT + marketing 🌴 West coast vs east coast culture
1
2
16
5,372
Mario Moscatiello retweeted
Placekey 2.0 is live! Placekey is an open entity resolution for places and addresses. Placekey helps with deduping, matching, linking, syncing, and merging physical places.
6
8
1
70
6,735
Mario Moscatiello retweeted
16
20
9
192
44,554