craenius
Is There a Way to Access Segment Anything or Sam-HQ via an API?
My system isn't the best to run inferences on models like this and would love to see if anyone is providing API access to these models.
Re:Hiring: Founding Engineer, Founding Designer
on Friday November 17, 2023 @05:43PM (#38311607)
oops. There's no way to delete this now, is there?
Hiring: Founding Engineer, Founding Designer
I am building ColdBook - an AI-driven emailing platform to help startup founders and sales folks.Currently going through ideation and MVP stage.Email: [email protected] to get involved.
Re:Building the Internet for AI
on Wednesday October 25, 2023 @12:10PM (#38014277)
What do you mean RAG as a service here? Is it just getting internet access to AI models? If so aren't companies like forefront/perplexity doing the same?
Re:Ask HN: Who wants to be hired? (October 2023)
on Wednesday October 04, 2023 @03:58AM (#37762183)
Location: India
Remote: yes
Willing to relocate: In 5 months
Technologies: Python, C++, Rust, business intelligence, deep learning. Founder of one startup, founding Backend engineer of another
Resume: https://drive.google.com/file/d/1GHd7nwyDaWfaQ9Qbe8Rp4jDp0QK...
Email: [email protected]
Founded first startup at age 14, founding Backend engineer of another at age 20. Developed NeuralDistance: method of predicting distance for ADAS and robotics application based on monocular images and neural networks. ML engineer since 2019 with a focus on NLP and business intelligence
What's a startup you would like to see in the AI space?
Is there a problem that's been bugging you for so long and no one seems to be addressing it?
Re:Any startups working on RAG infra for LLMs?
on Tuesday October 03, 2023 @02:15AM (#37748495)
Anyone you know of working on privacy preserving RAG pipelines or API access to contextual information?
Any startups working on RAG infra for LLMs?
Isn't on demand RAG an important area to work on inorder to enable realtime access and contextual understanding to LLMs? Something that can index the web and serve it using semantic tags or something? Is there a point to pursuing this, or anyone working on it or is this a dead end?