Tim Carambat

4.5K posts

Tim Carambat banner
Tim Carambat

Tim Carambat

@tcarambat

Building @AnythingLLM

Southern California Katılım Kasım 2018
1.1K Takip Edilen3.2K Takipçiler
Tim Carambat retweetledi
AI at AMD
AI at AMD@AIatAMD·
We want to extend a huge thank you to our #AIDevDay speakers and workshop hosts! 🤝 We are committed to building with you. @danielhanchen, @UnslothAI @JunchenJiang, @tensormesh @RyanLeeMiniMax @BanghuaZ, @thu_yushengsu, @RadixArk @clattner_llvm, @Modular @Stanford CS students Natalia Pahlavan, Laasya Konidala, Annmaria Antony @TCarambat, @AnythingLLM @gaeldelalleau, Augustin Verneuil, @Kog__AI Kevin Cochrane, @Vultr @QuentinAnthon15, @ZyphraAI
AI at AMD tweet mediaAI at AMD tweet mediaAI at AMD tweet mediaAI at AMD tweet media
English
7
10
41
5.6K
Tim Carambat retweetledi
IBM Developer
IBM Developer@IBMDeveloper·
🧵Building AI apps rarely means using just one model. Granite 4.1 brings language, vision, speech, and guardrails together—so you can build real workflows, not just demos. What devs should know.👇
English
3
17
72
13.6K
Tim Carambat retweetledi
Sahin Lale
Sahin Lale@SahinLale·
Check out how you can use Ternary Bonsai 8B 🌳 for tool calling in your everyday life—an impressive demo on an amazing platform by @AnythingLLM and @tcarambat!
English
1
8
19
2.2K
Tim Carambat
Tim Carambat@tcarambat·
@bindureddy "trust us, we did not train on the bench! The bench says we have to honor GPL and no lab has ever violated a copyright law!"
Tim Carambat tweet mediaTim Carambat tweet media
English
0
0
0
449
Bindu Reddy
Bindu Reddy@bindureddy·
INSANELY-BIG is a very large model from a frontier AI lab It scores 99 on literally EVERY SINGLE EVAL except SWE Bench Pro On SWE-BENCH pro it score 99.99 They say they won’t release it - cause it costs $100 per output 1M tokens and yeah - it’s too scary to drop to the general public
English
53
15
518
106.1K
Tim Carambat
Tim Carambat@tcarambat·
x.com/tcarambat/stat… Having done thing within AnythingLLM for a while now and easily do more than 10K meetings a day across our userbase. Totally on device with no cloud across Mac & Windows. All meetings transcripts and such are on device so agent tooling can also read them directly
Tim Carambat@tcarambat

Now is probably a great time to mention @AnythingLLM has shipped with an "open Granola" alternative for a while now. - Speaker Identification - Doesn't join meetings. Records speakers + host - any app - Actually fully on device (no cloud at all) - Custom summary templates - Mac & Windows (x64 and ARM) - All meetings indexed and searchable (locally) - Chat with transcripts (also local) Totally free, also does like 40 other things on top of meetings. You can also record in real time AND just upload arbitrary audio files. youtu.be/TrM1FzKrz5I

English
0
0
0
99
Idea Browser
Idea Browser@ideabrowser·
Startup idea that customers are begging to buy right now. Open Granola -Same meeting capture (no bot joining calls) -macOS Native desktop app -Notes stored as markdown files. -API (not MCP) -Local-first, or privacy. - Agent-readable by default. -Plays nicely with ai-first tools. Granola just locked down their local db, went MCP only. Someone is going to build a more open version and print money. If you build this... 1. I'll buy it and be your first customer. 2. I'll help you launch and get users. DM me the link.
Guido Appenzeller@appenz

Sorry to see Granola @meetgranola going closed. They encrypted their local db, no local and no cloud API. In a world where notes are managed by agents, the app now has zero value. Any recommendations for good alternatives? What are you switching to?

English
71
7
244
49.6K
Tim Carambat
Tim Carambat@tcarambat·
@omarwasm x.com/tcarambat/stat…
Tim Carambat@tcarambat

Now is probably a great time to mention @AnythingLLM has shipped with an "open Granola" alternative for a while now. - Speaker Identification - Doesn't join meetings. Records speakers + host - any app - Actually fully on device (no cloud at all) - Custom summary templates - Mac & Windows (x64 and ARM) - All meetings indexed and searchable (locally) - Chat with transcripts (also local) Totally free, also does like 40 other things on top of meetings. You can also record in real time AND just upload arbitrary audio files. youtu.be/TrM1FzKrz5I

QME
0
0
0
138
Omar Waseem
Omar Waseem@omarwasm·
Who’s launching OpenGranola
English
17
2
62
10K
Tim Carambat
Tim Carambat@tcarambat·
Now is probably a great time to mention @AnythingLLM has shipped with an "open Granola" alternative for a while now. - Speaker Identification - Doesn't join meetings. Records speakers + host - any app - Actually fully on device (no cloud at all) - Custom summary templates - Mac & Windows (x64 and ARM) - All meetings indexed and searchable (locally) - Chat with transcripts (also local) Totally free, also does like 40 other things on top of meetings. You can also record in real time AND just upload arbitrary audio files. youtu.be/TrM1FzKrz5I
YouTube video
YouTube
English
4
1
9
806
Tim Carambat
Tim Carambat@tcarambat·
@corbin_braun They build the apps first so that they can later release a model for people to build those same apps competing with the apps they built because....? If this were true, they would just never release it and paywall it for millions/mth B2B.
English
0
0
1
734
corbin
corbin@corbin_braun·
I want you to notice something about Anthropic. They seem to be dropping all these new products in their suite super fast. you are asking why? because I am 100% convinced they have created the end game AI model. whatever they call it, whether that is "Opus 6" etc. They have the model, that they know once they release it. People will be able to build real apps, in days. they want to grab as much of the market as possible before they do.
English
366
134
3.9K
387.2K
Tim Carambat
Tim Carambat@tcarambat·
@imrogb It's using whatever model you set! It is not fixed to a specific model. Open the settings > Ai providers > LLM and change to whatever provider and model you wish, local or cloud
English
0
0
0
56
Roger
Roger@imrogb·
@tcarambat tried the new meeting assistant as a stealth alternative to Fathom, liked it a lot but the Summary sucked, if i run the transcript through Claude it nails it, what LLM is the Summary using and how can we improve? Thanks
English
1
0
0
79
Tim Carambat retweetledi
🪱
🪱@worm_emoji·
Also hilarious how you see people talk about their local LLM setup for their personal data that sends all the data to a hosted provider. Missing the point entirely
English
7
5
351
28.3K
Tim Carambat
Tim Carambat@tcarambat·
This isn't just a transcript. Because it lives in AnythingLLM, it’s has access to all other tooling like on-device search, agent tools, MCP's you installed, and more. Every meeting has post-meeting agentic items queued up and ready to run with a click. For example, you might have a JIRA MCP queue up some tickets after an engineering meeting. Your LLM of choice is doing the summarization and making this determination, like it would with a traditional agentic chat.
English
1
0
2
536
Tim Carambat
Tim Carambat@tcarambat·
We just added a full AI Meeting Assistant into @AnythingLLM Desktop. It’s 100% local. It’s free. And it processes a 3-hour meeting in 3 minutes. It's always fun to nerd-snipe multi-million-dollar valued companies into a single feature We want you to cancel your Granola/Otter/Fireflies subscriptions. 🧵👇
English
8
2
17
1.6K
Tim Carambat
Tim Carambat@tcarambat·
@DylanEMHolland @basedjensen The instance only runs for the duration of a meeting. So an hour average. Almost certainly using spot because there was a comment about fulfillment. So assume a t3.small for example at 0.007/hr = 126k/month in just the instances
English
0
0
1
42
Dylan Holland
Dylan Holland@DylanEMHolland·
@basedjensen Latest funding round was 38 million, where the cheapest tier ec2 is ~$3 a month so 18m ec2 instances ~= 55m dollars a month
English
1
0
2
170
Hensen Juang
Hensen Juang@basedjensen·
I am sorry But 18 million instances ? I have run some of the largest clusters in the world serving 100s of millions of users and we were no where close to 18m nodes
David Gu@davidgu

we run 18 million EC2 instances per month. At our scale, we see very rare bugs very frequently. Last week, we received *half* an HTTP request. Not a HTTP 206, literally half a request. Content-Length was 2350 bytes. Body was actually 1200 bytes, and was truncated mid json doc.

English
69
14
1.6K
359.4K

Keşfet

@danielhanchen @UnslothAI @JunchenJiang @tensormesh @RyanLeeMiniMax @BanghuaZ @thu_yushengsu @RadixArk