Pinned Tweet
Diving into the New AIOps tools landscape: Unpacking fully assembled AI agents built on @evidence_dev, with data gathering and orchestration by @crewAIInc.
Why I did it:
🔍 To test whether agents can accurately gather comprehensive company info from a name and URL.
💼 My day job revolves around developing AI solutions that need constant monitoring and evaluation, so I explored the current offerings.
🌐 A quick Google search mostly shows giants like New Relic and Datadog. I wanted to highlight emerging startups in the space.
I've got a detailed market overview pending fact-checking. I compiled it over a few evenings, squeezing it in after my main projects. Stay tuned!
I think @cloudflare should follow @vercel's example and support sponsorship of such OSS projects
Introducing Localflare 🔥
The missing dashboard for local @Cloudflare development
One command. Zero config. Full visibility into your D1, KV, R2, Queues, and Durable Objects.
Here's what I built ↓
Built w/ @shadcn @vite_js @honojs
@dok2001 @elithrar @threepointone @ritakozlov @boristane @theo @CloudflareDev @burcs @eastdakota @KankaniAbhishek @irvinebroque
Sean Dorofeev retweeted
Memory v2 is now the default memory driver starting v1.4 🚀
Multi-user, Multi-session, Multi-Agent 🧡
🚀 Launching Memory 2.0: A multi-user, multi-session, chatgpt-like memory driver for Agents.
Memory v2 has been in beta for a few weeks and today it becomes the default Memory driver for all @AgnoAgi Agents. Read more 👇
I've been developing agents for over a year, and these guys are going super right. Top 3 best frameworks now. 👍🔥🚀
This is a perfect example of how to increase adoption into your product. my opinion is that this should be an absolute base for all AIs. Bravo @mastra
Introducing @mastra’s MCP documentation server: turn any AI assistant or IDE that has MCP support ( @cursor_ai, @windsurf_ai, @cline, etc ) into an expert on Mastra’s complete knowledge base.
Sean Dorofeev retweeted
GPT-4.5 is ready!
good news: it is the first model that feels like talking to a thoughtful person to me. i have had several moments where i've sat back in my chair and been astonished at getting actually good advice from an AI.
bad news: it is a giant, expensive model. we really wanted to launch it to plus and pro at the same time, but we've been growing a lot and are out of GPUs. we will add tens of thousands of GPUs next week and roll it out to the plus tier then. (hundreds of thousands coming soon, and i'm pretty sure y'all will use every one we can rack up.)
this isn't how we want to operate, but it's hard to perfectly predict growth surges that lead to GPU shortages.
a heads up: this isn’t a reasoning model and won’t crush benchmarks. it’s a different kind of intelligence and there’s a magic to it i haven’t felt before. really excited for people to try it!
Main Challenges of the AI Agent:
- Deploying it into production
- The use of various approaches leading to inconsistency and incompleteness
However, the Langgraph Supervisor could address all these issues simultaneously. This is a significant development in the CrewAI space.
github.com/langchain-ai/lang…
Sean Dorofeev retweeted
twitter hype is out of control again.
we are not gonna deploy AGI next month, nor have we built it.
we have some very cool stuff for you but pls chill and cut your expectations 100x!
Building AI products requires a new playbook. Don't solve problems that'll be solved by next-gen models.
Go narrow, focus on context, and don't worry about costs - today's Ferraris are tomorrow's Hondas.
Forget everything you know about building startups. AI products need a different playbook.
High costs and rapid model improvements mean you should focus on being exceptional at one thing, rather than trying to build for scale.
This might be @crewAIInc's most awaited commit, finally squashing those bugs that drive me nuts!
Have you seen this? My dev team looks like they need a spa day... or a one-way ticket to a tropical island!
Thank goodness for @modal_labs
Deploying in under five? That's like spotting a unicorn at the pizza shop: total magic!
That’s quite ironic! I’ll let myself make one more joke about @crewAIInc; I think I’ve earned it. The easiest and, more importantly, the fastest way to get OSS CrewAI into the cloud is to become a client of the corporate giant IBM 😅
P.S. The course is actually good; I recommend it!
Clarkson's Farm best illustration of startup life, probably best illustration then silicon valley’s series.
