The best Hacker News stories from Show from the past day
Latest posts:
Show HN: Open-source ETL framework to sync data from SaaS tools to vector stores
Hey hacker news, we launched a few weeks ago as a GPT-powered chatbot for developer docs, and quickly realized that the value of what we’re doing isn’t the chatbot itself. Rather, it’s the time we save developers by automating the extraction of data from their SaaS tools (Github, Zendesk, Salesforce, etc) and helping transform it to contextually relevant chunks that fit into GPT’s context window.<p>A lot of companies are building prototypes with GPT right now and they’re all using some combination of Langchain/Llama Index + Weaviate/Pinecone + GPT3.5/GPT4 as their stack for retrieval augmented generation (RAG). This works great for prototypes, but what we learned was that as you scale your RAG app to more users and ingest more sources of content, it becomes a real pain to manage your data pipelines.<p>For example, if you want to ingest your developer docs, process it into chunks of <500 tokens, and add those chunks to a vector store, you can build a prototype with Langchain fairly quickly. However, if you want to deploy it to customers like we did for BentoML ([<a href="https://www.bentoml.com/](https://www.bentoml.com/)" rel="nofollow">https://www.bentoml.com/](https://www.bentoml.com/)</a>) you’ll quickly realize that a naive chunking method that splits by character/token leads to poor results, and that “delete and re-vectorize everything” when the source docs change doesn’t scale as a data synchronization strategy.<p>We took the code we used to build chatbots for our early customers and turned it into an open source framework to rapidly build new data Connectors and Chunkers. This way developers can use community built Connectors and Chunkers to start running vector searches on data from any source in a matter of minutes, or write their own in a matter of hours.<p>Here’s a video demo: [<a href="https://youtu.be/I2V3Cu8L6wk](https://youtu.be/I2V3Cu8L6wk)" rel="nofollow">https://youtu.be/I2V3Cu8L6wk](https://youtu.be/I2V3Cu8L6wk)</a><p>The repo has instructions on how to get started and set up API endpoints to load, chunk, and vectorize data quickly. Right now it only works with websites and Github repos, but we’ll be adding Zendesk, Google Drive, and Confluence integrations soon too.
Show HN: Walkie-talkie for teams
Hey team, I'm Arjun! Builder at flowy.live.<p>So - it's quite simple. It's a walkie-talkie with a way to play back the last 24 hours. A new medium of communication for active teams.<p>We have released for Mac and Windows.<p>Any feedback or thoughts would mean the world to us!<p>Open to a feedback session? <a href="https://calendly.com/arjun-flowy/onboarding" rel="nofollow">https://calendly.com/arjun-flowy/onboarding</a><p>Appreciate all of you,
Arjun Patel
arjun@flowy.live
Show HN: Marvin – build AI functions that use an LLM as a runtime
Hey HN! We're excited to share our new open-source project, Marvin. Marvin is a high-level library for building AI-powered software. We developed it to address the challenges of integrating LLMs into more traditional applications. One of the biggest issues is the fact that LLMs only deal with strings (and conversational strings at that), so using them to process structured data is especially difficult.<p>Marvin introduces a new concept called AI Functions. These look and feel just like regular Python functions: you provide typed inputs, outputs, and docstrings. However, instead of relying on traditional source code, AI functions use LLMs like GPT-4 as a sort of “runtime” to generate outputs on-demand, based on the provided inputs and other details. The results are then parsed and converted back into native data types.<p>This “functional prompt engineering” means you can seamlessly integrate AI functions with your existing codebase. You can chain them together with other functions to form sophisticated, AI-enabled pipelines. They’re particularly useful for tasks that are simple to describe yet challenging to code, such as entity extraction, semantic scraping, complex filtering, template-based data generation, and categorization. For example, you could extract terms from a contract as JSON, scrape websites for quotes that support an idea, or build a list of questions from a customer support request. All of these would yield structured data that you could immediately start to process.<p>We initially created Marvin to tackle broad internal use cases in customer service and knowledge synthesis. AI Functions are just a piece of that, but have proven to be even more effective than we anticipated, and have quickly become one of our favorite features! We’re eager for you to try them out for yourself.<p>We’d love to hear your thoughts, feedback, and any creative ways you could use Marvin in your own projects. Let’s discuss in the comments!
Show HN: Marvin – build AI functions that use an LLM as a runtime
Hey HN! We're excited to share our new open-source project, Marvin. Marvin is a high-level library for building AI-powered software. We developed it to address the challenges of integrating LLMs into more traditional applications. One of the biggest issues is the fact that LLMs only deal with strings (and conversational strings at that), so using them to process structured data is especially difficult.<p>Marvin introduces a new concept called AI Functions. These look and feel just like regular Python functions: you provide typed inputs, outputs, and docstrings. However, instead of relying on traditional source code, AI functions use LLMs like GPT-4 as a sort of “runtime” to generate outputs on-demand, based on the provided inputs and other details. The results are then parsed and converted back into native data types.<p>This “functional prompt engineering” means you can seamlessly integrate AI functions with your existing codebase. You can chain them together with other functions to form sophisticated, AI-enabled pipelines. They’re particularly useful for tasks that are simple to describe yet challenging to code, such as entity extraction, semantic scraping, complex filtering, template-based data generation, and categorization. For example, you could extract terms from a contract as JSON, scrape websites for quotes that support an idea, or build a list of questions from a customer support request. All of these would yield structured data that you could immediately start to process.<p>We initially created Marvin to tackle broad internal use cases in customer service and knowledge synthesis. AI Functions are just a piece of that, but have proven to be even more effective than we anticipated, and have quickly become one of our favorite features! We’re eager for you to try them out for yourself.<p>We’d love to hear your thoughts, feedback, and any creative ways you could use Marvin in your own projects. Let’s discuss in the comments!
Show HN: RoboPianist, a piano playing robot simulation in the browser
Show HN: RoboPianist, a piano playing robot simulation in the browser
Show HN: RoboPianist, a piano playing robot simulation in the browser
Show HN: YakGPT – A locally running, hands-free ChatGPT UI
Greetings!<p>YakGPT is a simple, frontend-only, ChatGPT UI you can use to either chat normally, or, more excitingly, use your mic + OpenAI's Whisper API to chat hands-free.<p>Some features:<p>* A few fun characters pre-installed<p>* No tracking or analytics, OpenAI is the only thing it calls out to<p>* Optimized for mobile use via hands-free mode and cross-platform compressed audio recording<p>* Your API key and chat history are stored in browser local storage only<p>* Open-source, you can either use the deployed version at Vercel, or run it locally<p>Planned features:<p>* Integrate Eleven Labs & other TTS services to enable full hands-free conversation<p>* Implement LangChain and/or plugins<p>* Integrate more ASR services that allow for streaming<p>Source code: <a href="https://github.com/yakGPT/yakGPT">https://github.com/yakGPT/yakGPT</a><p>I’d love for you to try it out and hear your feedback!
Show HN: YakGPT – A locally running, hands-free ChatGPT UI
Greetings!<p>YakGPT is a simple, frontend-only, ChatGPT UI you can use to either chat normally, or, more excitingly, use your mic + OpenAI's Whisper API to chat hands-free.<p>Some features:<p>* A few fun characters pre-installed<p>* No tracking or analytics, OpenAI is the only thing it calls out to<p>* Optimized for mobile use via hands-free mode and cross-platform compressed audio recording<p>* Your API key and chat history are stored in browser local storage only<p>* Open-source, you can either use the deployed version at Vercel, or run it locally<p>Planned features:<p>* Integrate Eleven Labs & other TTS services to enable full hands-free conversation<p>* Implement LangChain and/or plugins<p>* Integrate more ASR services that allow for streaming<p>Source code: <a href="https://github.com/yakGPT/yakGPT">https://github.com/yakGPT/yakGPT</a><p>I’d love for you to try it out and hear your feedback!
Show HN: YakGPT – A locally running, hands-free ChatGPT UI
Greetings!<p>YakGPT is a simple, frontend-only, ChatGPT UI you can use to either chat normally, or, more excitingly, use your mic + OpenAI's Whisper API to chat hands-free.<p>Some features:<p>* A few fun characters pre-installed<p>* No tracking or analytics, OpenAI is the only thing it calls out to<p>* Optimized for mobile use via hands-free mode and cross-platform compressed audio recording<p>* Your API key and chat history are stored in browser local storage only<p>* Open-source, you can either use the deployed version at Vercel, or run it locally<p>Planned features:<p>* Integrate Eleven Labs & other TTS services to enable full hands-free conversation<p>* Implement LangChain and/or plugins<p>* Integrate more ASR services that allow for streaming<p>Source code: <a href="https://github.com/yakGPT/yakGPT">https://github.com/yakGPT/yakGPT</a><p>I’d love for you to try it out and hear your feedback!
Show HN: Gut – An easy-to-use CLI for Git
Hi Hacker news !<p>I’m Julien and I built an alternative CLI for Git : gut.<p>Even if I haven’t been coding for a long time (I’m in the first year studying computer science), I’ve always found git to be frustrating.
The command naming is inconsistent and git lets you easily shoot yourself in the foot.<p>I made gut, another git porcelain, to solve these issues.<p>It provides a consistent naming of command. To do so, syntax is based on subcommands. For example, to delete a branch, run gut branch rm rather than git branch -d, same to delete a remote (gut remote rm) and so on.<p>Gut also prevents you from shooting yourself. It provides nice defaults and always prompt you before doing something destructive.
Also, it won’t allow you to rewrite the history if it has been pushed to the remote. Creating commits in detached head is also prohibited.<p>Finally, git was made when GitHub and others didn’t existed yet. To diff commits, gut opens the compare view in the browser. And to merge a branch, gut opens a pull request.<p>I have been working on this project for the past few months and I am happy to be able to share it.<p>I hope you’ll like it. Any suggestions is welcome !<p>Repo: <a href="https://github.com/julien040/gut">https://github.com/julien040/gut</a>
Show HN: Gut – An easy-to-use CLI for Git
Hi Hacker news !<p>I’m Julien and I built an alternative CLI for Git : gut.<p>Even if I haven’t been coding for a long time (I’m in the first year studying computer science), I’ve always found git to be frustrating.
The command naming is inconsistent and git lets you easily shoot yourself in the foot.<p>I made gut, another git porcelain, to solve these issues.<p>It provides a consistent naming of command. To do so, syntax is based on subcommands. For example, to delete a branch, run gut branch rm rather than git branch -d, same to delete a remote (gut remote rm) and so on.<p>Gut also prevents you from shooting yourself. It provides nice defaults and always prompt you before doing something destructive.
Also, it won’t allow you to rewrite the history if it has been pushed to the remote. Creating commits in detached head is also prohibited.<p>Finally, git was made when GitHub and others didn’t existed yet. To diff commits, gut opens the compare view in the browser. And to merge a branch, gut opens a pull request.<p>I have been working on this project for the past few months and I am happy to be able to share it.<p>I hope you’ll like it. Any suggestions is welcome !<p>Repo: <a href="https://github.com/julien040/gut">https://github.com/julien040/gut</a>
Show HN: StratusGFX, my open-source real-time 3D rendering engine
It's been closed source for a long time while I worked on it on and off as a hobby research project, but yesterday the repo was made public for the first time under the MPL 2.0 license.<p>A feature reel showing its capabilities can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/feature_reel" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/feature_re...</a><p>A technical breakdown of a single frame can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/frame_analysis" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/frame_anal...</a><p>It's still in a very beta state (bugs and instability expected), but I felt like it was a good time to make it public since a lot of its core features are mostly presentable. I plan to continue working on it in my spare time to try and improve the usability of the code.<p>Two main use cases I could see for it:<p>1) People using it for educational purposes.<p>2) People integrating it into other more general purpose engines that they're working on since Stratus is primarily a rendering engine. Any extensions to the rendering code that are made public would then further help others.<p>So I think it will remain very niche but I'm hoping it will still be helpful for people in the future.
Show HN: StratusGFX, my open-source real-time 3D rendering engine
It's been closed source for a long time while I worked on it on and off as a hobby research project, but yesterday the repo was made public for the first time under the MPL 2.0 license.<p>A feature reel showing its capabilities can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/feature_reel" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/feature_re...</a><p>A technical breakdown of a single frame can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/frame_analysis" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/frame_anal...</a><p>It's still in a very beta state (bugs and instability expected), but I felt like it was a good time to make it public since a lot of its core features are mostly presentable. I plan to continue working on it in my spare time to try and improve the usability of the code.<p>Two main use cases I could see for it:<p>1) People using it for educational purposes.<p>2) People integrating it into other more general purpose engines that they're working on since Stratus is primarily a rendering engine. Any extensions to the rendering code that are made public would then further help others.<p>So I think it will remain very niche but I'm hoping it will still be helpful for people in the future.
Show HN: StratusGFX, my open-source real-time 3D rendering engine
It's been closed source for a long time while I worked on it on and off as a hobby research project, but yesterday the repo was made public for the first time under the MPL 2.0 license.<p>A feature reel showing its capabilities can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/feature_reel" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/feature_re...</a><p>A technical breakdown of a single frame can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/frame_analysis" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/frame_anal...</a><p>It's still in a very beta state (bugs and instability expected), but I felt like it was a good time to make it public since a lot of its core features are mostly presentable. I plan to continue working on it in my spare time to try and improve the usability of the code.<p>Two main use cases I could see for it:<p>1) People using it for educational purposes.<p>2) People integrating it into other more general purpose engines that they're working on since Stratus is primarily a rendering engine. Any extensions to the rendering code that are made public would then further help others.<p>So I think it will remain very niche but I'm hoping it will still be helpful for people in the future.
Show HN: StratusGFX, my open-source real-time 3D rendering engine
It's been closed source for a long time while I worked on it on and off as a hobby research project, but yesterday the repo was made public for the first time under the MPL 2.0 license.<p>A feature reel showing its capabilities can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/feature_reel" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/feature_re...</a><p>A technical breakdown of a single frame can be found here:
<a href="https://ktstephano.github.io/rendering/stratusgfx/frame_analysis" rel="nofollow">https://ktstephano.github.io/rendering/stratusgfx/frame_anal...</a><p>It's still in a very beta state (bugs and instability expected), but I felt like it was a good time to make it public since a lot of its core features are mostly presentable. I plan to continue working on it in my spare time to try and improve the usability of the code.<p>Two main use cases I could see for it:<p>1) People using it for educational purposes.<p>2) People integrating it into other more general purpose engines that they're working on since Stratus is primarily a rendering engine. Any extensions to the rendering code that are made public would then further help others.<p>So I think it will remain very niche but I'm hoping it will still be helpful for people in the future.
Show HN: Mirrorful – A developer-first way to implement designs faster
Hey HN! Mirrorful (<a href="https://www.mirrorful.com/">https://www.mirrorful.com/</a>) is an open-source developer framework that helps front-end engineers manage their design systems. We’ve been building Mirrorful with the open-source community (<a href="https://github.com/Mirrorful/mirrorful">https://github.com/Mirrorful/mirrorful</a>) and wanted to share our beta with you. Check out our online demo to get the idea: <a href="https://app.mirrorful.com/">https://app.mirrorful.com/</a>.<p>Design systems can be thought of as the “building blocks of your app” which makes me think of Lego bricks. Mirrorful helps you manage your codebase’s Lego bricks and ensure that they are consistent across all of your apps and platforms.<p>We saw as product engineers how hard it is to get code to match Figma mock ups. High-quality design is a competitive advantage, so getting your UI pixel perfect can matter a lot, but is time-consuming and tedious.<p>When we worked for large public companies, we saw that good component libraries help, but engineers are often still dealing with tweaking small design decisions. There are a lot of inefficiencies. We also worked at a small startup and saw what it was like to not have a design system. No design system led to copy pasta code, and days of back-and-forth on simple things like “what hex should i be using for the hover state?”<p>Design systems are tricky to get right. Picking an out-of-the-box solution is easy to begin with, but one day you’ll be cursing yourself due to lack of flexibility (we did!). On the other hand, creating a design system from scratch is super time-consuming even for the best frontend engineers. Mirrorful is our way out of this dilemma.<p>Mirrorful is completely open-source and written in Typescript. We’re starting with basic design elements—commonly called “design tokens” — such as colors, typography, and shadows, but have plans to expand our scope into more complex components.<p>As frontend engineers ourselves, we wanted a tool that lives in code but is visual. It had to be super easy to set up, but also prepare you for scale so you and/or your team don’t end up copy-pasting everywhere. We decided to make it an NPM package (<a href="https://www.npmjs.com/package/mirrorful" rel="nofollow">https://www.npmjs.com/package/mirrorful</a>) that runs a localhost editor and exports out your design tokens into any configuration you want: .js, .ts, .css, .scss, .json. It’s lightweight with no design system lock-in.<p>Our product is completely self-serve: just install our NPM package. If you run Mirrorful locally, a visual dashboard will pop up at localhost:5050 that lets you manage your theme and export various configuration files directly into code.<p>Pricing is similar to other open-source companies—we charge for cloud-hosted features and for premium components.<p>We’ve built open-source/open-core projects before and love interacting with contributors from all over the world. If anyone has any opinions on what we’re building, we’re all ears. Check us out at mirrorful.com and at github.com/Mirrorful/mirrorful and give it a shot!
Show HN: Mirrorful – A developer-first way to implement designs faster
Hey HN! Mirrorful (<a href="https://www.mirrorful.com/">https://www.mirrorful.com/</a>) is an open-source developer framework that helps front-end engineers manage their design systems. We’ve been building Mirrorful with the open-source community (<a href="https://github.com/Mirrorful/mirrorful">https://github.com/Mirrorful/mirrorful</a>) and wanted to share our beta with you. Check out our online demo to get the idea: <a href="https://app.mirrorful.com/">https://app.mirrorful.com/</a>.<p>Design systems can be thought of as the “building blocks of your app” which makes me think of Lego bricks. Mirrorful helps you manage your codebase’s Lego bricks and ensure that they are consistent across all of your apps and platforms.<p>We saw as product engineers how hard it is to get code to match Figma mock ups. High-quality design is a competitive advantage, so getting your UI pixel perfect can matter a lot, but is time-consuming and tedious.<p>When we worked for large public companies, we saw that good component libraries help, but engineers are often still dealing with tweaking small design decisions. There are a lot of inefficiencies. We also worked at a small startup and saw what it was like to not have a design system. No design system led to copy pasta code, and days of back-and-forth on simple things like “what hex should i be using for the hover state?”<p>Design systems are tricky to get right. Picking an out-of-the-box solution is easy to begin with, but one day you’ll be cursing yourself due to lack of flexibility (we did!). On the other hand, creating a design system from scratch is super time-consuming even for the best frontend engineers. Mirrorful is our way out of this dilemma.<p>Mirrorful is completely open-source and written in Typescript. We’re starting with basic design elements—commonly called “design tokens” — such as colors, typography, and shadows, but have plans to expand our scope into more complex components.<p>As frontend engineers ourselves, we wanted a tool that lives in code but is visual. It had to be super easy to set up, but also prepare you for scale so you and/or your team don’t end up copy-pasting everywhere. We decided to make it an NPM package (<a href="https://www.npmjs.com/package/mirrorful" rel="nofollow">https://www.npmjs.com/package/mirrorful</a>) that runs a localhost editor and exports out your design tokens into any configuration you want: .js, .ts, .css, .scss, .json. It’s lightweight with no design system lock-in.<p>Our product is completely self-serve: just install our NPM package. If you run Mirrorful locally, a visual dashboard will pop up at localhost:5050 that lets you manage your theme and export various configuration files directly into code.<p>Pricing is similar to other open-source companies—we charge for cloud-hosted features and for premium components.<p>We’ve built open-source/open-core projects before and love interacting with contributors from all over the world. If anyone has any opinions on what we’re building, we’re all ears. Check us out at mirrorful.com and at github.com/Mirrorful/mirrorful and give it a shot!
Show HN: Mirrorful – A developer-first way to implement designs faster
Hey HN! Mirrorful (<a href="https://www.mirrorful.com/">https://www.mirrorful.com/</a>) is an open-source developer framework that helps front-end engineers manage their design systems. We’ve been building Mirrorful with the open-source community (<a href="https://github.com/Mirrorful/mirrorful">https://github.com/Mirrorful/mirrorful</a>) and wanted to share our beta with you. Check out our online demo to get the idea: <a href="https://app.mirrorful.com/">https://app.mirrorful.com/</a>.<p>Design systems can be thought of as the “building blocks of your app” which makes me think of Lego bricks. Mirrorful helps you manage your codebase’s Lego bricks and ensure that they are consistent across all of your apps and platforms.<p>We saw as product engineers how hard it is to get code to match Figma mock ups. High-quality design is a competitive advantage, so getting your UI pixel perfect can matter a lot, but is time-consuming and tedious.<p>When we worked for large public companies, we saw that good component libraries help, but engineers are often still dealing with tweaking small design decisions. There are a lot of inefficiencies. We also worked at a small startup and saw what it was like to not have a design system. No design system led to copy pasta code, and days of back-and-forth on simple things like “what hex should i be using for the hover state?”<p>Design systems are tricky to get right. Picking an out-of-the-box solution is easy to begin with, but one day you’ll be cursing yourself due to lack of flexibility (we did!). On the other hand, creating a design system from scratch is super time-consuming even for the best frontend engineers. Mirrorful is our way out of this dilemma.<p>Mirrorful is completely open-source and written in Typescript. We’re starting with basic design elements—commonly called “design tokens” — such as colors, typography, and shadows, but have plans to expand our scope into more complex components.<p>As frontend engineers ourselves, we wanted a tool that lives in code but is visual. It had to be super easy to set up, but also prepare you for scale so you and/or your team don’t end up copy-pasting everywhere. We decided to make it an NPM package (<a href="https://www.npmjs.com/package/mirrorful" rel="nofollow">https://www.npmjs.com/package/mirrorful</a>) that runs a localhost editor and exports out your design tokens into any configuration you want: .js, .ts, .css, .scss, .json. It’s lightweight with no design system lock-in.<p>Our product is completely self-serve: just install our NPM package. If you run Mirrorful locally, a visual dashboard will pop up at localhost:5050 that lets you manage your theme and export various configuration files directly into code.<p>Pricing is similar to other open-source companies—we charge for cloud-hosted features and for premium components.<p>We’ve built open-source/open-core projects before and love interacting with contributors from all over the world. If anyone has any opinions on what we’re building, we’re all ears. Check us out at mirrorful.com and at github.com/Mirrorful/mirrorful and give it a shot!
Show HN: go-nbd – A Pure Go NBD Server and Client
Hey HN! I just released go-nbd, a lightweight Go library for effortlessly creating NBD servers and clients. Its a neat tool for creating custom Linux block devices with arbitrary backends, such as a file, byte slice or what I'm planning to use it for, a tape drive. While there are a few partially abandoned projects like this out there already, this library tries to be as maintainable as possible by only implementing the most recent handshake revision and baseline functionality for both the client and the server, while still having enough support to be useful.<p>I'd love to get your feedback :)