On-device AI development assistant for your entire workflow
Start new thread
Kevin William David

Pieces Copilot+ - Remember anything with a real-time, on-device AI assistant

by•
Ask questions about anything on your desktop computer. With live context from your entire workflow, your personal AI copilot can summarize online research, resolve complex coding issues, and streamline every aspect of your work-in-progress journey.

Add a comment

Replies

Best
Keith Mauppa
I’ve been using Pieces copilot for several months now and it’s mind-blowing! ❤️🤟🏾 I'm not a trained developer, but I can code a little and I create solutions for my team (in my spare time). Pieces has been indispensable for my coding projects. It's the first window I open whenever I'm inspired to start coding something. I have OpenAI, Claude, Gemini, etc. and I make use of these tools as well, but Pieces is in a different league. It's purpose-built for developer workflows by developers and it shows! I started using Pieces in like Q3 '23 as my Coding Snippet Library and then they added the Live Context Co-pilot earlier this yr (long before Microsoft's failure with Recall) 😏 P.S. Live Context is AIR GAPPED for those of you worried about Privacy. And Pieces rocks even without this feature imo. In fact to be critical, I think the Live Context feature (while awesome) has slowed down the performance of Pieces for me slightly and made it a more resource intensive application overall - not as resource intensive as VSCode or PyCharm though. I wonder if the Live Context should be an optional add-on or separate app? 🤔
Rosie Levy
Maker
@keimaumau thank you SO MUCH for being an early adopter and for sticking with us this long!! That is a good point - like you mentioned, Pieces requires more resources than it used to due to the fact that we are running everything on-device for privacy and security purposes. You can disable the Workstream Pattern Engine (the magic behind Live Context) at any time to reduce the load. However, the release we sent out last night should have improved this a decent amount, so please try it out today and let me know what you think!
Andreas Sohns
I love how this platform focuses on real time data processing right on the device. It's perfect for developers who value privacy.
Rosie Levy
Maker
@andreas_sohns we are so glad you think so, being local first is so integral to our company and our goals!
Daniel
Do I understand right? It is an AI that watches everything I do on the screen and can help me based on what it sees me doing?
Rosie Levy
Maker
@daniel_d7 essentially yes it can, but only what you allow it to see and entirely on-device! If you enable the Workstream Pattern Engine, it will be able to extract key information using a vision processing stack, but only from what you have on your screen. All of this is captured, processed, and stored entirely on-device, and then the Pieces Copilot can leverage this information to give you seriously helpful responses! For more details, check out this article: https://code.pieces.app/blog/an-...
Daniel
@rosieatpieces It's a really nice idea! Isn't Microsoft is doing something similar for Windows 11 ?
Rosie Levy
Maker
@daniel_d7 Yes, some other companies are hoping to achieve a similar goal in being able to extract segmentations of the foreground application, however they are not leveraging this in the highly contextual and workflow-specific way we are, and they are most certainly not doing it entirely on-device like we do!
Daniel
@rosieatpieces Ok thanks for the explanation. Is there any price for that? I can't find any mention of that
Rosie Levy
Maker
@daniel_d7 We are offering Pieces for free at the moment as we work to build an amazing product, not just for individual developers but soon for teams and enterprise. We will begin offering PiecesPro later this year and have posted a draft pricing plan here: https://github.com/pieces-app/su... Would love your feedback on this!!
Ashik Hameed
@tsavo_at_pieces Just checked out Pieces Copilot+ and it’s amazing! The real-time, on-device AI assistant is exactly what I've been looking for. The ability to ask questions about anything on my desktop and get live context from my workflow is a great. It's so helpful for summarizing online research, resolving complex coding issues, and streamlining every aspect of my work. The privacy aspect is also a big plus. I love the specific prompts you mentioned, like summarizing pull request comments, recalling key takeaways from past research, and coordinating team syncs. It sounds like this will seriously boost productivity. Looking forward to diving in and seeing how much it improves my workflow. Kudos to you and the team for such a fantastic product!
Shivay Lamba
Maker
@tsavo_at_pieces @ashikhameed Thanks a lot for your feedback! Do let us know if you have any questions when trying out Pieces
Kehui Guo
Wow!! Congrats on the launch! Looks like Pieces is integrating all of my most used platforms, love how seamless the experience it. Good luck!
Mr. Ånand
Thanks for the support! @kehui_guo Looking forward to see your use-cases across your tool chain :)
Rosie Levy
Maker
@kehui_guo we are trying our best to be a truly horizontal and completely integrated workflow assistant! Thank you for the kind words!
Andrew Baisden
Congrats and good luck on the launch @tsavo_at_pieces and Pieces team! Pieces is one of my favourite tools right now and I am getting so much value from it. I have already seen my productivity increase a lot!
Rosie Levy
Maker
@tsavo_at_pieces @andrewbaisden thank you so much for the support!!
Stefans Keiss
I'm curious about the on-device AI aspect. How does it handle privacy and data security?
Shivay Lamba
Maker
@memphys_sk Pieces Copilot+ with Live Context ensures your data remains secure and private by leveraging on-device processing and intelligent data handling techniques. This approach provides a secure, efficient, and privacy-focused tool for enhancing developer productivity. Here are the main features: Privacy and Security at the Core:On-Device Processing: All workflow data is processed and stored locally on your device, ensuring sensitive information never leaves your machine unless explicitly shared. Intelligent Visual Snapshots: Instead of continuous screen recording, PiecesOS captures intelligently timed snapshots of application visuals, processed on-device to ensure only relevant information is analyzed. Temporary Data Handling: Extracted text is temporarily stored in memory and deleted after processing, typically within 100 milliseconds. Irrelevant data is discarded after 12 hours. Data Compression and Redaction: Relevant data is compressed to about 10% of its original size using an on-device transformer model, filtering out sensitive information like API keys and PII. Secure Local Storage: Summarized data is stored in a local vector database on your device, queried during retrieval-augmented generation (RAG) sessions without transmitting data to the cloud. Optional Cloud Integration: You can use cloud-based LLMs if preferred, but only refined context is sent to the cloud. On-device LLM runtimes ensure maximum privacy. Secure Integrations: For integrations like the VS Code Plugin, data exchanges remain secure via local HTTP/GRPC connections.
Meni Teknikio
This is a great move in the world of AI.
Mark Widman
Maker
@meni_teknikio Thanks for your support! Let us know what you think about the product!
Kyrylo Silin
Hey Tsavo, Having an on-device AI assistant to streamline work sounds incredibly productive. How does the AI handle privacy and data security given that it processes workflow data in real-time? Also, what are some of the key use cases you envision for this tool? Congrats on the launch!
Sam Jones
Maker
Hey there @kyrylosilin Key here is that our architecture is offline first. You can choose from 8+ local LLMs (my favorite right now is LLAMA 3) to handle all of the processing locally. You can leverage cloud LLMs if you'd like but our objective is to give users domain and control over their data. You can check out this page for more information on privacy and security: https://docs.pieces.app/product-... As for use cases, the sky is the limit. Personally I love: - generating github issues from emails and chat - fixing and iterating on code without having to copy paste what i want to work on into chat - summarizing research papers and finding common themes in a body of research Would love to hear more of your thoughts as you experiment!
Sinikiwe Phahlane
Congrats on the launch! 🎉 The ability to ask questions about anything on your desktop and get real-time context from your workflow is incredibly powerful. Can't wait to see how it helps with online research, coding issues, and more! 🚀💻
Shivay Lamba
Maker
@sinikiwe_phahlane Thanks a lot for your response. Glad you find it useful!
First
Previous
123
•••
Next
Last