Knapsack

Knapsack

Technology, Information and Internet

San Francisco, California 101 followers

AI you can take to work.

About us

Creators of Knapsack, which provides fast, on-device AI. Carve your own path with a secure, on-premise, open-source GPT. Semantic search and chat with ultimate privacy and security. Available initially on Mac, with PC coming soon. Register to get on the waitlist.

Website
https://knap.ai
Industry
Technology, Information and Internet
Company size
2-10 employees
Headquarters
San Francisco, California
Type
Privately Held
Founded
2023

Locations

Employees at Knapsack

Updates

  • View organization page for Knapsack, graphic

    101 followers

    A great talk by Cooper previewing our application at GitHub.

    View profile for Curtis Barker, graphic

    Experienced Product Leader

    How to oprationalize GAI beyond the 50 shades of internet chat bots? Adding color to this debate, i had the pleasure to host a great line up of builders and an energetic audience at GitHub offices in San Francisco this week for another AICamp meetup event. ▶ Christy Bergman, Dev Advocate from Zilliz demonstrated using the latest AI models with proprietary data on a customizable RAG stack. This is a key motion for businesses looking to opertionalize AI. ▶ Cooper Lindsey , CEO of Knapsack demonstrated how to deploy large and small learning models locally to overcome AI cost and data security concerns. ▶ Finally Michael Vandi , Founder of Addy AI and Spatika Ganesh showcased thier AI-powered email that aims to deliver the next gen email experience we all need! 🙏 Huge thanks GitHub for supporting the AI developers community, and the amazing AICamp bay area team on-site Devinder Sodhi, and behind the scenes Sonam G. , Hamza Zerarka Elisa Yao, for bringing these events together. 💪🏽 📣 The event recording is available at AICamp. Check us out on Slack and Discord and look out for the next AICamp meet ups, workshops and hackathons coming soon! #llm,#llms, #aicommunity, #ai, #genain

    • No alternative text description for this image
    • No alternative text description for this image
    • No alternative text description for this image
    • No alternative text description for this image
  • View organization page for Knapsack, graphic

    101 followers

    View profile for Cooper Lindsey, graphic

    Founder and CEO, Knap / Building Knapsack, the 100% private AI app.

    Some news: my co-founder Mark Heynen and I have agreed to take over the Stanford AI Startups Meetup group -- please sign up! I teamed up with Mark initially on PayJoy right after graduating from Stanford in 2016 and again in the last year to start Knap, a company focused on fast on-device/on-premise AI services. Mark and I will be bringing more events to you in the coming months. We have been active in AI Camp and other well-organized groups, and with all the excitement and interest in AI, we think there is so much potential in these IRL events. Please let us know if there are talks you want to give or see. https://lnkd.in/gKcH4ijy

    Stanford AI Startups | Meetup

    Stanford AI Startups | Meetup

    meetup.com

  • View organization page for Knapsack, graphic

    101 followers

    View profile for Mark Heynen, graphic

    Ex Google, Meta, and 4x founder. Advisor, operator and investor focused on AI, fintech, and emerging markets.

    Cooper Lindsey and I have decided to bring on a senior software engineer to help us pioneer how people can use open source LLMs to streamline their work at Knap. The ideal candidate will have a strong foundation in programming languages such as Python, Rust, or Typescript. While these skills are beneficial, they are not essential requirements. What is essential is a combination of intelligence, speed, and the ability to work independently, and at least three years of relevant experience, preferably with an undergraduate degree in Computer Science. We are looking for someone who can take high ownership of the full stack of our AI applications (i.e., front end and back end). We value a positive and collaborative work ethic, with no room for egos. The ideal candidate will also be comfortable in an early stage company, with either a portion of their payment deferred until we raise our next round or an equity component.  On time allocation, we are open to part time, e.g.,  20 hours per week.  We are also open to contract work if that is preferred. While not essential, candidates with experience in vectorization, vector databases, deploying open-source LLMs, or developing desktop applications will be highly valued. We are excited to find the right individual to join our team and contribute to our mission of making AI safe and easy for people to use at work. If you are interested in this opportunity, please reach out.

  • View organization page for Knapsack, graphic

    101 followers

    The myth of closed-source superiority is fading in the world of LLMs. Open Source LLMs are catching up, with some surpassing in specific tasks.  Check out these advancements: Jukebox: Open-source music generation model, rivaling commercial options.  Megatron-Turing NLG: Open-source text generation model, generating human-quality news articles.  Goliath 120B:  A developer took two Llama-70B models, stitched them together, and it performed well. Openness fosters innovation, leading to better models for everyone. Let's ride the open source wave! https://lnkd.in/gaCp8zrx  

    ChatGPT Is One-Year Old: Are Open-Source Large Language Models Catching Up?

    ChatGPT Is One-Year Old: Are Open-Source Large Language Models Catching Up?

    cobusgreyling.medium.com

  • View organization page for Knapsack, graphic

    101 followers

    Imagine a world where just a few companies control the #AI tools we use daily. That's the squeeze we face with big tech deployments, which creates risks: - Limited innovation:  New features and ideas gates by large companies and their roadmaps - Opacity: Black box algorithms, bias concerns unaddressed. - Vendor lock-in:  Switching costs soar, freedom to choose evaporates. - Shifting data sources: Depending on the outcome of lawsuits, training data quality can vary - Changing terms: API terms and requirements can change on a moments’ notice - High costs: When a model learns and improves, these learnings cannot be accessed easily by others. Open Source LLMs offer a powerful alternative: transparency, collaboration, and democratization of AI. It's time to break the squeeze! 🐙  We have been digging into this, and will be sharing more in the coming days.

    • No alternative text description for this image

Similar pages