Product Thumbnail

Instill VDP

Open source unstructured data ETL for AI first applications

Open Source
Developer Tools
Artificial Intelligence
GitHub
No-Code
Data

Versatile Data Pipeline (VDP): An open-source, no-code/low-code solution for quick AI workflow creation. It handles unstructured data, ensuring efficient data connections, flexible pipelines, and smooth AI model and data source integration.

Top comment

Instill's drag and drop interface makes it very easy to build complex AI workflows interacting with many models (OpenAI, HuggingFace, ...). And it ingest many data types (text, image, audio, ...) Really powerful for anybody that needs to build things visually in nocode.

Comment highlights

This open-source AI workflow solution is a game-changer! It's incredibly easy to use and allows for quick creation of AI workflows without the need for coding. The handling of unstructured data is seamless, and the integration with AI models and data sources is smooth. This is a must-have tool for anyone working with AI.
If you're into AI and data, you need to check out Instill VDP on Product Hunt. It's redefining data workflow efficiency
This will be the future that everyone can create AI applications with easy-to-use no-code/low-code tools. Thanks for making this happen and being transparent with open-source strategy.
Instill VDP strikes the perfect balance between sophistication and simplicity in how to handle unstructured data. Excited to see its impact on our AI projects.
Congratulations to the team on the launch! You're doing great (and there are many of you there). Every day, the processes are becoming easier. I haven't delved into the product yet, but what I've seen is really cool. Where can I find implementations of complex cases (complex for me) for language models?
I really love Instill's service, very easy to use. I also love the vision of keeping it open-sourced. This is what we need in the AI era!
Very talented team and super focused! Wish Instill AI the best on product launch!
Amazing Instill VDP tool! Very cool part for creating a no-code or low-code tool.
WoW~ Instill VDP is reeeeeally revolutionizing the AI tech landscape with its effortless handling of complex data! It’s not just an innovation, it’s also a game-changer. If you’re looking for a tool that simplifies complexity, your search ends here. Give it a try and experience the future of AI!
My experience with Instill VDP, especially in integrating pipelines from GPT API to Stable Diffusion, has been very smooth and user-friendly. Highly recommended.
Hello, Product Hunt community! 👋 I’m Xiaofei, the founder and COO of Instill AI. Thrilled to introduce Instill Versatile Data Pipeline (VDP) here. Thanks to @picsoung for featuring Instill VDP, and kudos to our developer advocate, @shih_chun_huang! 💧 What Instill VDP is (and isn’t) Back in 2022, when we first introduced VDP, we made it clear that Instill VDP is not your typical ML platform catering to a specific vertical, nor does it exclusively focus on a single MLOps component. Today, I’d like to add another aspect to this “What VDP isn’t” list—it’s not just another LLM app building tool. VDP’s scope isn’t limited to LLM or solely text data. Our motivation behind Instill VDP stemmed from our frustration with the available tool stacks for MLOps related to unstructured data, including text, images, videos, audio, and more. We aimed to create a general, modular, and flexible ETL framework tailored for unstructured data. ⭐ Why open-source We’re committed to avoiding further fragmentation in the MLOps and AI tool landscape. Unstructured data ETL is a significant challenge, and our approach is all about developing an open and extensible framework. We encourage various integrations and aim to empower communities to benefit from and actively participate in the process. 👐 Why Instill Hub I believe many of you appreciate the value of code-sharing and model-sharing platforms like GitHub and Hugging Face. But what about sharing our approaches to handling unstructured data? Do we really need to reinvent the wheel every time we build workflows or data pipelines from scratch? Instill Hub serves as a platform to explore, create, share, clone, and customize data pipelines effortlessly. You can also obtain production-ready APIs that can seamlessly integrate into your codebase. 👉 We value your feedback and invite you to join our community to help refine the ETL infrastructure for unstructured data. https://discord.gg/sevxWsqpGh
Hey everyone! Founder and CEO at Instill AI here. Thank you @picsoung for featuring Instill VDP (https://github.com/instill-ai/vdp). It’s our very first time here. We’re super excited! Big thanks to @shih_chun_huang for showcasing the beta features in the demo. I’d like to share a bit more about what drives our product and the passion of the team. Our aim with Instill VDP is to create an awesome no-code/low-code tool that lets you work with all kinds of data - whether it’s unstructured, semi-structured, or structured, and from any source - be it public cloud, private cloud, or on-premise. The cool part? You can either host it yourself, bring your own cloud (BYOC), or use it on our public Instill Cloud. We’ve tried to keep the pricing for VDP on the Instill Cloud as wallet-friendly as possible. This all stems from our mission to make AI accessible to everyone. We’re also super keen to see how you all use VDP, which is why we created Instill Hub. Think of it like GitHub for sharing code, or Hugging Face for sharing models, but Instill Hub for sharing VDPs. While GPT Store has millions of GPT apps, Instill Hub is your go-to for finding versatile data pipelines – like a RAG-based ChatBot, vehicle detection with YOLOv7, a speech-to-text assistant, and more. I’m here to answer any questions and really appreciate your feedback. For more details on our launch, check out our announcement article: https://www.instill.tech/blog/vd...