Ravi Parikh

28 posts

Ravi Parikh banner
Ravi Parikh

Ravi Parikh

@ravisparikh

Katılım Kasım 2009
8 Takip Edilen4.9K Takipçiler
Ravi Parikh retweetledi
pink
pink@iJaadee·
Amazing. Canva, but automatic. Girls you’ll want to get on this. Here’s a quick marketing hack 🌸—> Give Moda a PDF / Slideshow / Etc Media of your product. Ask it to create an animated, marketable presentation based on it 🌸—> Have an AI of your choice do the research to collect contact information for local business, etc. potential clients. 🌸—> Use Claude to create an email template, tell it to customize the template based on the company it’s emailing, and send a minimum of 50 emails per day using the Moda presentation as the pitch (aka, “why you should hire me!”) This is the new 2026 way of applying of for jobs. With more agency and creative liberty. ••• 🩷••• If you’re extra skilled, you can even Run this chain through a loop where the Moda pitch is customized each time for the specific pain points of the client company being emailed. Customization = higher customer acquisition
Anvisha@anvisha

We raised $7.5M to kill AI slop. Introducing Moda: the world's first design agent with taste. RT+ comment “Moda” and we’ll design your brand for FREE.

English
4
7
132
8.6K
Steve Ruiz
Steve Ruiz@steveruizok·
canvas testing on @trymoda @anvisha. (cool product, not trashing, canvas is hard, etc etc) 1. Good job having pointer capture (to allow for dragging outside of the parent container) but need to turn off user select on the UI at least when in an interaction state
English
5
1
37
3.1K
Ravi Parikh
Ravi Parikh@ravisparikh·
@steveruizok the clamped resize is by design (image nodes i.e. rects with pattern fill don't let you resize beyond the bounds of the image unless it's repeating, similar to canva UX). The bug where it doesn't resize properly when you go into negative resize is a bug, but not rotation-related
English
1
0
2
45
Steve Ruiz
Steve Ruiz@steveruizok·
Rotation is hell and will fuck with all other systems. Sadly you'll be squashing bugs forever unless you really design your layout system around transforms from the beginning.
English
3
0
2
994
Ravi Parikh
Ravi Parikh@ravisparikh·
@hwchase17 deep agents has been huge for us. I managed to delete 16k lines of code when we integrated it into our agent architecture
English
1
0
3
121
Ravi Parikh retweetledi
Harrison Chase
Harrison Chase@hwchase17·
congrats to Moda on the launch of their design agent! it's built on deep agents - all agents going forward should be built on these types of agent harnesses if youre still using the same agent architecture from 1.5 years ago, time to re-evaluate blog: blog.langchain.com/how-moda-build…
Anvisha@anvisha

We raised $7.5M to kill AI slop. Introducing Moda: the world's first design agent with taste. RT+ comment “Moda” and we’ll design your brand for FREE.

English
13
11
129
21.7K
Anvisha
Anvisha@anvisha·
We raised $7.5M to kill AI slop. Introducing Moda: the world's first design agent with taste. RT+ comment “Moda” and we’ll design your brand for FREE.
English
2.5K
1.8K
7.8K
4.3M
Ravi Parikh
Ravi Parikh@ravisparikh·
I’ve tried every AI slide maker at this point. The experience is always the same: you type a prompt, something comes back, and it looks so far from anything you’d actually ship that you end up starting from scratch. The models aren’t the problem. The problem is that most of these tools are built on abstractions that were never meant for AI. They’re either writing HTML or operating on scene models that don’t expose the right structure for layout and visual reasoning. Moda rebuilt the canvas from the ground up on WebGPU. The agent has first-class access to objects, hierarchy, bounds, grouping, alignment, spacing, constraints, and composition. It has a much richer model of the canvas than systems that design through a translation layer. You can feel the difference the first time you use it.
Anvisha@anvisha

We raised $7.5M to kill AI slop. Introducing Moda: the world's first design agent with taste. RT+ comment “Moda” and we’ll design your brand for FREE.

English
2
0
23
2.3K
Ravi Parikh retweetledi
Datapoint AI
Datapoint AI@datapointai·
1/ Introducing Datapoint AI Intelligence is being commoditized. Good judgement is still rare. We collect human preference data that trains frontier models, by replacing mobile ads with preference tasks.
English
4
16
33
8.8K
Ravi Parikh retweetledi
Anvisha
Anvisha@anvisha·
Logo grids used to be 30 mins of googling SVGs and aligning boxes. Now it's one prompt
English
4
2
28
5.5K
Ravi Parikh retweetledi
Anvisha
Anvisha@anvisha·
What if the "🍌 Beautify this slide" button in Google slides actually worked
English
1
5
30
7.5K
Ravi Parikh retweetledi
Anvisha
Anvisha@anvisha·
GPT 5.4 vs Opus 4.6 for Powerpoint @claudeai still wins but it's closer than before - kind of a matter of taste at this point. GPT 5.4: + more interesting layouts - more mistakes - took 3 min per slide (!) - cost $2.76 Opus: - less layout diversity + more consistent, polished without iterations + took <1 min per slide + cost $0.98 BTW I'm using @trymoda agent harness for slides, which gets you way better outputs than raw pptx or Google Slides. Opus does more with less, but GPT 5.4 has its place. Still got more tests to run.
Anvisha tweet media
English
2
4
29
10.6K
Ravi Parikh retweetledi
Anvisha
Anvisha@anvisha·
Love waking up to this kind of user feedback
Anvisha tweet media
English
0
1
5
1.2K
Ravi Parikh retweetledi
Stefano Ermon
Stefano Ermon@StefanoErmon·
Mercury 2 is live 🚀🚀 The world’s first reasoning diffusion LLM, delivering 5x faster performance than leading speed-optimized LLMs. Watching the team turn years of research into a real product never gets old, and I’m incredibly proud of what we’ve built. We’re just getting started on what diffusion can do for language.
English
320
588
4.2K
982.2K
Ravi Parikh retweetledi
Dan Robinson
Dan Robinson@danlovesproofs·
First "AI for slides" product I've been excited about. The AI is manipulating the same canvas you are, so it feels more like cursor and less like reprompting midjourney 20x per slide.
Anvisha@anvisha

Finally, AI designs you can edit. Moda - create brand assets on a fully editable canvas. This is not another ChatGPT or nano banana wrapper. We actually taught an agent layout, typography, and color. We gave it taste. Use it for slides, social, infographics, e-books …

English
1
2
11
2.3K