Michael Seibel

NLP inside your database — Query OpenAI’s GPT-3 from directly inside your database

353
You’ve explored the power of GPT-3 & ChatGPT; now you can apply that power to your own data by bringing GPT-3 to your database with MindsDB, to deliver additional insights & value to your existing data. MindsDB is an Open-Source ML Platform for Developers

Add a comment

Replies
Best
Adam Carrigan
Hey Product Hunters! Adam here, co-founder of MindsDB, the fastest growing open source applied machine learning platform in the world 🙂 Today we are excited to launch our NLP integration with Open AI’s GPT-3 and Huggingface pre-trained models, bringing the power of two of the world's most popular NLP frameworks to your database. This means you can now harness the power of these models on your own data directly inside your existing database with only a few queries. We believe that all applications will & should be enhanced or powered by AI, but right now, this is painful and slow. The workhorse that is the full-stack developer sets up the database, builds the front end and calls for the API, but yet if they want to incorporate even the simplest Machine Learning into the process this requires a team of specialists, one that may not even be available in some companies. Thus at MindsDB, we want to empower the full-stack developer to add a new pillar to their tool kit….. Machine Learning, and with MindsDB this is now possible. MindsDB is an Open-Source Machine Learning Platform for Full Stack Developers. We currently integrate ~70 popular databases with state-of-the-art machine learning frameworks, and this number is constantly growing thanks to the amazing open-source community! So, here’s some of the things you can do (with just a few SQL commands 😲) all in your DB: Classify and label rich text, for instance, for sentiment analysis, detecting hate speech, or spam; Extract meaning for labelling text even when you don't have any training data - so-called zero-shot classification; Have conversations, answer questions or comments; Automatically summarize long texts, translate them, and more. Upcoming Webinar: We are also hosting a deep dive webinar on our NLP capabilities, including GPT-3 and Huggingface. Sign-up here Check out this Demo Video and Get Started with MindsDB. We would love your feedback. Kind regards, Adam
Adam Carrigan
Thank you everyone for your support today. We look forward to continuing the conversation in our community.
deepu
Congrats on the launch. For someone like me who is proficient in building UI and use Supabase as backend for basic CRUD where do I start in mindsdb?
Adam Carrigan
@pradeeb28 You can either deploy it on your own servers (like AWS) or you can start by testing it out on our cloud (https://cloud.mindsdb.com/home). We also have a quickstart on our docs here which is a great place to start. (https://docs.mindsdb.com/quickstart), and if you need any help our community is always around for support. (https://mindsdb.com/joincommunity)
Zoran Pandovski
@pradeeb28 Also, you will be able to integrate MindsDB directly with Supabase https://docs.mindsdb.com/data-in....
Jorge Torres
@pradeeb28 we are also working on SDKs, if you join our community on slack we can talk about the current SDK implementation.
Rutam Prita Mishra
@pradeeb28 If you're proficient in building UI and use Supabase as a backend for basic CRUD, getting started with MindsDB should be straightforward. Here are the steps you could follow to get started: 1. Sign up for MindsDB: You can sign up for a free account on the MindsDB website to start using the platform. 2. Explore the documentation: MindsDB provides comprehensive documentation that explains how to use the platform, including how to train models, make predictions, and access other features. 3. Connect to your data: MindsDB supports multiple data sources, including SQL databases like Supabase. You can connect to your Supabase database and import your data into MindsDB to start training models. 4. Train a model: Once you have your data in MindsDB, you can start training models by specifying the target column and selecting the features you want to use. MindsDB will automatically select the best model for your data and train it for you. 5. Make predictions: After your model is trained, you can start making predictions by sending new data to MindsDB. You can use the MindsDB API or the JavaScript SDK to make predictions from your front-end application. 6. Evaluate the output: Finally, you can evaluate the output of your models to ensure that they are accurate and make reasonable predictions. MindsDB provides tools for post-prediction analysis that allow you to examine the output and make sure that it meets your needs. Overall, the process of getting started with MindsDB is straightforward, and the platform provides many tools and features to help you train models and make accurate predictions quickly and easily.
George Hosu
Great to see you guys keep improving and getting more traction month after month, keep at it 😊
Zoran Pandovski
@george_hosu Thanks for all of your amazing contributions 🙌
Adam Carrigan
Thanks @george_hosu, we miss you :) hopefully catchup soon.
Ameed Jamous
Very promising project with cloud and community version 💪. We are already looking into it. Deserves an upvote.
Adam Carrigan
@telecomsxchange Thank you, our team is available to help should you have any questions.
Zoran Pandovski
@telecomsxchange Thanks. We are happy to help you in your ML Journey
Paolo De Giglio 🚀
Congrats on the launch of MindsDB on Product Hunt! 🎉 It looks like an amazing platform for full-stack developers to integrate machine learning into their applications. What types of use cases will the NLP integration with Open AI's GPT-3 and Huggingface pre-trained models enable? I'm excited to see what the possibilities are! 🤩 Keep up the great work!
Adam Carrigan
Hey @paolo_degiglio, Thanks for your question & support. We plan to support the majority of the Huggingface models and a large % of the ones from OpenAI. A list of all the things we currently support for Huggingface can be found here: https://docs.mindsdb.com/nlp/nlp... and for OpenAI here: https://docs.mindsdb.com/nlp/nlp.... As a short list, we support - Spam detection - Sentiment analysis - Emotion classifier - Hate speech detector -Question Detection But we also support many more.
Rutam Prita Mishra
@paolo_degiglio The NLP integration of MindsDB with OpenAI's GPT-3 and Hugging Face pre-trained models enables a wide range of use cases across different industries and applications. Here are a few examples: 1. Chatbots: MindsDB can be used to build advanced chatbots that can understand natural language queries and provide human-like responses. The integration with GPT-3 and Hugging Face models allows chatbots to have better language understanding and generate more accurate and relevant responses. 2. Sentiment Analysis: MindsDB can be used to build sentiment analysis models that can analyze and classify the sentiment of texts, such as customer reviews or social media posts. The integration with pre-trained NLP models enables sentiment analysis models to have better accuracy and provide more insightful results. 3. Text Classification: MindsDB can be used to build text classification models that can categorize texts into different classes, such as spam or ham, news articles, or product reviews. The integration with GPT-3 and Hugging Face models enables text classification models to have better accuracy and provide more relevant results. 4. Named Entity Recognition: MindsDB can be used to build named entity recognition models that can extract entities, such as names of people, organizations, and locations, from texts. The integration with GPT-3 and Hugging Face models enables named entity recognition models to have better accuracy and provide more relevant results. 5. Content Generation: MindsDB can be used to build content generation models that can generate new text based on given inputs. The integration with GPT-3 and Hugging Face models allows content generation models to have better language understanding and generate more coherent and relevant outputs. These are just a few examples of the types of use cases that can be enabled with the NLP integration of MindsDB with GPT-3 and Hugging Face models. The platform provides a user-friendly interface and a wide range of tools and features to help you build, train, and deploy NLP models for your specific use case, making it accessible to full-stack developers and non-experts alike.
Johannes
Very cool, will check this out! :)
Adam Carrigan
@jobenjada Thank you.
Martyna Slawinska
@jobenjada You can create a demo MindsDB account here: https://cloud.mindsdb.com/ Also, feel free to join our Slack community (https://mindsdb.com/joincommunity) and discover MindsDB through our docs (https://docs.mindsdb.com/)!
Rutam Prita Mishra
@jobenjada Great thing! Do let us know how this goes.
Macgill Davis
I love MindsDB and can't recommend the service more highly. We have used MindsDB Cloud to integrate ML within our product effortlessly. The team is incredibly supportive and helpful too. Congrats on the launch!
Adam Carrigan
@macgridcupcake Thanks, Macgill; awesome to be working with you. Props to the whole MindsDB team.
Zoran Pandovski
@macgridcupcake Thank you, it is a pleasure to work with you and the team.
Ianu82
@macgridcupcake thanks Macgill - we love Rize too :-)
Han Wang
Congrats on the launch! MindsDB is nothing short of a world class team and product. No better time than now to build something for folks in the LLM space 😉
Adam Carrigan
@han_wang6 Thankyou Han for your kind words.
Zoran Pandovski
Thank you @han_wang6. It's pleasure to work with you and the Mintlify team
Richard Gao
Do you have plans to connect image generation APIs as well to databases, or is it just NLP for now? Mainly asking because we have image generation with evoke-app.com for now, but we are planning to get some NLP models in the future. Congrats on your launch!
Richie Rambal
@richard_gao2 We are tackling Data from structured data onwards, I can see it happening in the near future, stay tunned !
Patricio Cerda-Mardini
@richard_gao2 @ricardo_r_f_ Yes! We're exploring API integrations at the moment to enable seamless interplay between them and MindsDB-exposed ML models, so definitely something we want to add in the short to medium term.
Rutam Prita Mishra
@richard_gao2 Thanks for your support. Hopefully, MindsDB Team can come up with Image Generation APIs connections as well soon to suit your needs.
Richard Gao
@ricardo_r_f_ @paxcema Good to hear! We provide an image generation (and soon to be general) AI API with Evoke and hope to get it integrated into MindsDB. Feel free to DM me on twitter if you have more questions about Evoke: https://twitter.com/TheRealEtch
Gurkirat Singh
Hey @adam_carrigan congrats on the launch. I would give it a try in my new side-product.
Adam Carrigan
@gurkirat_singh Thank you, if you'd like any support you can join our community. Our team are available to provide support. https://mindsdb.com/joincommunity
Richie Rambal
@adam_carrigan @gurkirat_singh Documentation is also pretty cool. https://docs.mindsdb.com
Konstantin Sivakov
Since I have used MindsDB for the last few months and have integrated it with CKAN, this will be an extremely useful feature. The MindsDB team has done an excellent job
Ianu82
@tino097 oh, interesting! What is the ML problem you’re solving with this please?
Konstantin Sivakov
@ianu82 Im working as CKAN (ckan.org) developer and ive created an adapter in MindsDB to use CKAN Datastore API as source for training models. However there was idea to use MindsDB libs to create an extension for CKAN for some users requremets/needs.
Ianu82
@tino097 Ah, thanks for the information Konstantin. Nice work!
Andrii Doroshev
Great, congratulations on your launch! Good luck to you. I'm already using GPT-3 in SEO.
Ianu82
@new_user_273dccefac wow, sounds cool - how are you using GPT-3 for SEO please?
Alejandro Villegas
@new_user_273dccefac @ianu82 I think on keywords research, workflows on SEO, Product descriptions ETC.
Dimitar Cetelev
Awesome feature ! :) What is your plan with the futures AI tools like from Google, Facebooks etc.. ? Are you going to integrate them all ?
Adam Carrigan
@dimitar_cetelev Absolutely; it just so happens Google's Bard launched today on Product Hunt as well. Our team is looking at how we can integrate it. Ultimately we are driven by our customers, users and community. What we build next is based on what they want/need.
Richie Rambal
@dimitar_cetelev On the roadmap we also have in store some beautiful integrations such as Bringing your own models and other AutoML technologies
David Martin
@dimitar_cetelev @ricardo_r_f_ Great this is on the roadmap
Jorge Torres
@dimitar_cetelev would you like to build some of these integrations?
Rutam Prita Mishra
@dimitar_cetelev Thanks for the appreciation. With the growing use cases and amazing community around, we can expect to have these integrations pretty soon.
Muhammad Sarmad Hafeez
Amazing! Bring the power of NLP to your database with MindsDB. Can't wait to see how this impacts data insight and analytics! #MindsDB #NLPDatabase
PAMELA VELASQUEZ SALAZAR
Congrats to the developers of this interesting tool. Very useful and easy to use. I loved the UI, very clean.
Gianna Castellanos
Congrats to the team behind this tool! It's really impressive what this can do! Really like the UI!
Patricio Cerda-Mardini
@gianna_castellanos Thank you Gianna! Truly means a lot.
Francois Delporte
Ohhhh super interesting !! I'll follow ;) Upvoted
Adam Carrigan
Thank you @wawa_trkl; excited to hear your feedback.
Martyna Slawinska
@wawa_trkl Feel free to join our Slack community (https://mindsdb.com/joincommunity) and discover MindsDB through our docs (https://docs.mindsdb.com/)!
Ankur Singh
Congrats on the launch, this is useful!!
Adam Carrigan
@singh_ankur Thank you Ankur
Zoran Pandovski
Thanks, @singh_ankur. Feel free to join our community and learn more about MindsDB.
Rutam Prita Mishra
@singh_ankur Thank you. Feel free to give it a try.
Jeremy Galarce
Congrats on the launch of MindsDB! 🚀🎉 🚀🎉