Ricardo Sequerra Amram

3.9K posts

Ricardo Sequerra Amram banner
Ricardo Sequerra Amram

Ricardo Sequerra Amram

@ric0seq

Partner @pointninecap | @projecteurope_

Worldwide Katılım Nisan 2009
2.1K Takip Edilen4.4K Takipçiler
Ricardo Sequerra Amram retweetledi
Christoph Janz 🕊
You're a founder trying to make your team AI-native. Shoot red pills at skeptical engineers, dodge meetings, and watch out for the VP of Resistance firing blue pills back! Convert 70% before time runs out. Your converts join your army. zingy-blini-722c70.netlify.app PS: I've just built this tiny game (v0.09) in less than an hour (which is kind of crazy and sort of proves the point). LMK if you like it and I might spend some time on improving it!
Christoph Janz 🕊 tweet media
English
1
1
7
1.7K
Ricardo Sequerra Amram retweetledi
Attio
Attio@attio·
Ask Attio is now in Slack. DM or mention Attio in a channel. Ask anything about your customers and deals without leaving where you already work. Search through records, calls, emails, and conversation history in seconds.
English
6
14
64
16.8K
Christoph Janz 🕊
Christoph Janz 🕊@chrija·
In London on Wed/Thu. I know this question has become a bit of a meme ... but ... who should I meet?
English
29
2
70
10.8K
Ricardo Sequerra Amram retweetledi
Christoph Janz 🕊
Christoph Janz 🕊@chrija·
Excited to formally launch P9's Scout Program!
English
3
2
50
5.7K
Ricardo Sequerra Amram retweetledi
Nikolaus West
Nikolaus West@NikolausWest·
This is so cool!
Pablo Vela@pablovelagomez1

We have SLAM on the Robocap! 🎉 Visualized with @rerundotio Using @NVIDIAAIDev cuVSLAM for GPU-accelerated multicamera tracking. I basically wrote zero code myself and fully used Claude Code for this. It worked because I had so many existing examples to point to that it just wrote everything the way I would have. A few technical wins: 1. Used rattler-build from @prefix_dev to package the compiled cuVSLAM CUDA binaries, which made it SUPER easy to use across repos. This also means it works on the DGX Spark (ARM64) out of the box. 2. Zero-setup experience: git clone && pixi run track-robocap auto-downloads a 100MB dataset from HuggingFace and tracks frames. 3. Real-time 3D visualization with trajectories, landmarks, pose graphs, and video playback in Rerun. Still visual-only (not visual-inertial yet), and loop closure needs some debugging. Next steps are getting this into a Gradio interface, then into daggr, and extending it to work with other datasets from exoego-forge. The last piece I'm excited about: Rerun's RRD files now support layers for incremental data. Planning to build pipelines that go from raw sensor data → slam -> human pose → depth estimation → etc. Repo here: github.com/rerun-io/roboc…

English
2
4
48
7.9K
Ricardo Sequerra Amram
Ricardo Sequerra Amram@ric0seq·
So Claude now sources one cool company a day At the rate @chrija is building, we'll have our Claude-sourced investment very soon Imagine when it's fully Claude-led 😅
English
2
1
23
4.2K
Ricardo Sequerra Amram
Ricardo Sequerra Amram@ric0seq·
People don't understand the scale of @chrija 's token usage The wise SaaS fox is AI-pilled to the bone 😆
Christoph Janz 🕊@chrija

It's been a busy weekend for my coding agents. 1) Implemented various improvements to my new AI Command Center (see medium.com/point-nine-new…). Amongst others, the research skill will now try to apply learnings from a few 1000 hours of subscribed podcast transcripts (@AcquiredFM, @dwarkesh_sp, @InvestLikeBest, @twentyminutevc). 2) Built a little integration that lets us update our Webflow website by talking to a Slackbot. 3) Started building an app that aggregates bank accounts and assets and turns it into a dashboard (using FinAPI and CSV imports). 90% done, but as usual, the last 10% might take longer. ;-) 4) Played around with an idea I had ... use the "thinking" output of AIs to generate a real/near-time video ... didn't get too far with this yet. :) Has anyone built this?

English
5
0
15
4.7K
Ricardo Sequerra Amram retweetledi
scott belsky
scott belsky@scottbelsky·
the orchestration layer is the new interface layer. as we spend our day coordinating agent workflows (in a model agnostic fashion, local and cloud) and validating outputs (human in the loop, and resolving issues), the ultimate layer to own is where coordination takes place.
English
68
40
542
71.9K
Ricardo Sequerra Amram
Ricardo Sequerra Amram@ric0seq·
the AI revolution is being televised day by day, on X, podcasts, videos and streams (potentially) the largest shift in human society is also the most documented
English
1
0
1
189
Ricardo Sequerra Amram retweetledi
Nikolaus West
Nikolaus West@NikolausWest·
This is another incredible banger of a release with a bunch of features that have been in the works for a long time. Rerun has always been very code first, which I think has been a large part of our success so far. That said, users have long asked for more flexibility in deciding what and how to visualize on the fly in the viewer, and this release takes a huge step in that direction with the ability to do just that for time series plots. As usual there is a powerful system underneath that we will be rolling out more broadly in the future. Stay tuned. The second huge thing in this release is the ability to visualize recordings that are much bigger than what can fit in memory. This brings the most value currently to our data platform customers that can browse and manage large datasets of long multimodal recordings and jump around quickly in long timelines to understand their data. The team really went the extra mile to make it a super snappy and performant experience. This is particularly important right now in as robot models are getting better and more teams are pushing on long range reliability. A year ago most learning teams were recoding short episodes but as reliability improves we see teams move to day long recordings and handling that smoothly is super important to us.
Rerun@rerundotio

Rerun 0.30.0 is out and its 🔥🔥 This release brings significant improvements for handling of larger recordings, scalar plotting, MCAP support, and making it easier to extend the Viewer without forking. Let's dig in 👇

English
0
2
11
1.4K