Ciprian Nastase
@CipAIExplorer
🚀💥AI Innovator in #CustomerSupport 🤖 | Driving 80% ticket resolution via #AI | Sharing insights on #AIForSupport & tech transformations.
Next-level RAG🤯: - multiple agents - web searches - LLM per-user memory - basic RAG I envision this integrated in AI support, with the LLM being able to provide not only responses backed up by real data sources, but also fine-tuned to each customer. Awesome demo @ashpreetbedi !
Introducing `Personalized Agentic RAG` - where the LLM remembers key details about the user and automatically chooses the tool for RAG. We'll build: 👏 ChatGPT like memory ⚙️ Function calling 🧙 Multi-Agent orchestration code: git.new/personalization
I find the AI labs often lack a clear conception of how AI adoption will really happen. They want to build agents that will just "do work." Surprisingly, they increasingly can, but that work doesn't result in value unless they are integrated into the organization & its processes
In today's lesson of "Unexpected AI Uses":
no, i'm not building a quick MCP server to login and scrape my son's schools parent portal for missing assignments and using that in an agent i'm building with AgentKit...why do you ask?
Ironically, I don't think OpenAI trusts their own models enough. A key to GPTs was that you can just talk to the model and it would build a GPT for you. But agent builder is all about setting up variables & uploading bits of code & defining loops. Why not let the AI decide that?
Everyone talks about model alignment. Nobody talks about organizational misalignment.
Zero-person billion dollar companies. Early AGI-like breakthroughs. Viral Sora deepfakes. Sam Altman broke it all down in our exclusive DevDay conversation. Highlights: 0:00 Intro 0:43 DevDay's biggest launches 2:13 Agent Builder vs GPT Builder 3:43 The no-code revolution for…
No, OpenAI's Agent Builder did not kill n8n and Zapier. It's a great little agent builder no doubt. And quite powerful with MCP, file integrations, and state management out of the box. But it's a long way off from cannibalizing n8n's user base. n8n, Zapier, and their…
When every company says “AI-powered,” the only edge left is being AI-competent.
Stop parsing unstructured LLM output. Force structure from the start. Structured outputs eliminate 90% of "AI integration" bugs. Example: class CodeReview(BaseModel): issues: list[str] severity: str suggested_fix: str # Model returns valid JSON, guaranteed review…
Maybe some of the big problems with vibe coding are process problems, not AI problems...
this paper shows that at this point the problems with vibe coding are mostly human hallucinations are rare and barely noticeable in practice the real issue is development driven by instant gratification and weak qa, like skipping tests and relying on llms for verification
Truly revolutionary - and it works.
This is literally the future: "Students spend only two hours in the morning on science, math and reading, working at their own speed using personalized, AI-driven software. Adults in the classroom are called guides, not teachers, and earn six-figure salaries. Their job is to…
New on the Anthropic Engineering Blog: Most developers have heard of prompt engineering. But to get the most out of AI agents, you need context engineering. We explain how it works: anthropic.com/engineering/ef…
We're not building calculators. We're building thinking partners. The humanity part isn't optional.
Fast progress in training AI agents to interact with the world. Training on just 2,541 hours of Minecraft video, Google built an AI that runs on a single GPU & was able to mine diamonds offline (which takes an average of 24,000 clicks). The same approach may work for AI robots.
Cursor can now control your browser. Agent can take screenshots, improve UI, and debug client issues. Try our early preview with Sonnet 4.5.
A few months back, inspired by a great session by @edjgeek , we made changes to an AWS step function reducing its cost from $450 per invocation to around $1. If you're intrigued, check the first part of a two-part article I wrote below medium.com/chronicles-of-…
Great read ! 5 minutes read on how to slash AWS cost by @alaa_m_nassef
A few months back, inspired by a great session by @edjgeek , we made changes to an AWS step function reducing its cost from $450 per invocation to around $1. If you're intrigued, check the first part of a two-part article I wrote below medium.com/chronicles-of-…
Introducing Claude Sonnet 4.5—the best coding model in the world. It's the strongest model for building complex agents. It's the best model at using computers. And it shows substantial gains on tests of reasoning and math.
Am I weird if I think this isn't truly solving a problem, and will instead just drive prices higher so that merchants will cover the "small fee" they are charged by OpenAI for each transaction? Are we really that lazy that we can't check out/buy things ourselves anymore?
We’re introducing the Agentic Commerce Protocol, an open standard co-developed with @Stripe that enables programmatic commerce flows between users, AI agents, and businesses.
United States الاتجاهات
- 1. Chiefs 101K posts
- 2. Branch 29.3K posts
- 3. Mahomes 30.7K posts
- 4. #TNABoundForGlory 50K posts
- 5. #LoveCabin N/A
- 6. LaPorta 10K posts
- 7. Goff 13.3K posts
- 8. Bryce Miller 4,201 posts
- 9. Kelce 15.6K posts
- 10. #OnePride 6,283 posts
- 11. Dan Campbell 3,288 posts
- 12. #LaGranjaVIP 45.9K posts
- 13. #DETvsKC 4,810 posts
- 14. Butker 8,334 posts
- 15. Mariners 47.9K posts
- 16. Pacheco 4,870 posts
- 17. Gibbs 5,498 posts
- 18. Baker 53.7K posts
- 19. collinsworth 2,911 posts
- 20. Mike Santana 3,953 posts
Something went wrong.
Something went wrong.