The best Hacker News stories from Show from the past week

Go back

Latest posts:

Show HN: Matrix Multiplication with Half the Multiplications

Show HN: Matrix Multiplication with Half the Multiplications

Show HN: Skyvern – Browser automation using LLMs and computer vision

Hey HN, we're building Skyvern (<a href="https://www.skyvern.com">https://www.skyvern.com</a>), an open-source tool that uses LLMs and computer vision to help companies automate browser-based workflows. You can see some examples here: <a href="https://github.com/Skyvern-AI/skyvern#real-world-examples-of-skyvern">https://github.com/Skyvern-AI/skyvern#real-world-examples-of...</a> and there's a demo video at <a href="https://github.com/Skyvern-AI/skyvern#demo">https://github.com/Skyvern-AI/skyvern#demo</a>, along with some instructions on running it locally.<p>We provide a natural-language API to automate repetitive manual workflows that happen within the companies' backoffices. You can check out our code and play with Skyvern here: <a href="https://github.com/Skyvern-AI/Skyvern">https://github.com/Skyvern-AI/Skyvern</a><p>We talked to hundreds of companies about things they do in the background and found that most of them depend on repetitive manual workflows. The breadth of these workflows surprised us – most companies started off doing things manually, and eventually either hired people to scale the manual work, or wrote scripts using Selenium-like browser automation libraries.<p>In these conversations, one common point stood out: scaling is a pain either way. Companies relying on hiring struggled to adjust team sizes with fluctuating demand. Companies using Selenium and similar tools had a different problem: it can take days or even weeks to get a new workflow automated, and then would require ongoing maintenance any time the underlying websites changed because their XPath based interaction logic suddenly became invalid.<p>We felt like there was a way to get the best of both worlds with LLMs. We could use LLMs to reason through a website’s layout, while preserving the advantage of traditional browser automations allowing it to scale alongside demand. This led us to build Skyvern with a few core functionalities:<p>1. Skyvern can operate on websites it’s never seen before by connecting visible elements with the natural language instructions provided to us. We use a blend of computer vision and DOM parsing to identify a set of possible actions on a website, and multi-modal LLMs to map the natural language instructions to the available actions on the page.<p>2. Skyvern is resistant to website layout changes, as it doesn’t depend on any predetermined XPaths or other selectors. If a layout ever changes, we can leverage the methodology in #1 to complete the user-specified goal.<p>3. Skyvern accepts a blob of information when navigating workflows—basically just a json blob of whatever information you want to put, and then we use LLMs to map that to information on the screen. For example: if you're generating a quote from Geico, they commonly ask “Were you eligible to drive at 21?”. The answer could be inferred from the driver receiving their license in 2012, and having a birth date of 1996.<p>The above strategy adapts well to a number of use cases that Skyvern is helping companies with today: (1) Automating materials procurement by searching for, adding to cart, and transacting products through vendor websites that don’t have APIs; (2) Registering accounts, filing forms, and searching for information on government websites (ex: registering franchise tax information for Delaware C-corps); (3) Generating insurance quotes by completing multi-step dynamic forms on insurance websites; (4) Automating the job application process by mapping user-specified information (such as a Resume) to a job posting.<p>And here are some use-cases we’re actively looking to expand into: (1) Automating post-checkup data entry with patient data inside medical EHR systems (ie submitting billing codes, adding notes, etc), an (2) Doing customer research ahead of discovery calls by analyzing landing pages and other metadata about a specific business.<p>We’re still very early and would love to get your feedback!

Show HN: A user-friendly UI for viewing and editing Markdown files

Show HN: Flox 1.0 – Open-source dev env as code with Nix

Hey HN,<p>I'm Ron Efroni, CEO at Flox, and today we are releasing version 1.0 of our open source CLI, helping folks manage development environments everywhere.My own experience with development environments began with air-gapped systems, having to actually burn software to a CD to iterate over a very slow and expensive development cycle, sometimes reaching the server rack and realizing I have the wrong disk.... Fast forward to today and there are countless alternatives available backed by incredible compute resources, yet we somehow still find ourselves paying the price of long development cycles. That's why I've been working for over a decade to simplify the development stack so we can spend more time on making 1's and 0's do magical things, and why my co-founder Michael and I started Flox to bring you the solution based on Nix. Today is just the first step on that journey. We hope you'll take a peek at our new release, and very much look forward to continuing the journey with you from here together!<p>Introducing Flox 1.0<p>Flox is a platform that lets developers and operators focus on building fast with reproducible environments that span the enterprise SDLC. Using a declarative framework based on Nix, a package management and configuration tool, Flox allows developers to create environments that contain everything they need to build software.<p>Why Flox?<p>Flox behaves a lot like your favorite and familiar package manager, but it allows you to create as many environments as you want on your machine. Each one can contain a different combination of packages.<p>Environments are portable by default. If you install a package inside one that isn't cross-platform, it's easy to carve out exceptions. It's also easy to write hooks and populate your environment with variables - we designed it to be hackable.<p>Flox environments run in user-space, like, where you are. When you type `ls` after activating a Flox environment you will see the same stuff because you're in the same place - even with all those new packages available. No mounting volumes, no proxying ports. No breaking into the toolset you just conjured.<p>Getting Started: No sign-ups, just one install away. Dive into our GitHub repository (<a href="https://github.com/flox/flox">https://github.com/flox/flox</a>) and start exploring<p>I’m around all day to answer questions, talk Nix, or just reminisce about simpler times ;).<p>Lots of open source love, Ron

Show HN: Comma Separated Values (CSV) to Unicode Separated Values (USV)

Show HN: I made a free animator. Think Adobe Illustrator but for animation

Trangram is a free one-stop platform to create, and share motion graphics and svg animations with a free built-in powerful editor which is a fusion of Adobe Illustrator and animation tools.

Show HN: Teable – Open-Source No-Code Database Fusion of Postgres and Airtable

Features<p>Spreadsheet-like interface All you want is here • Cell Editing: Directly click and edit content within cells.<p>• Formula Support: Input mathematical and logical formulas to auto-calculate values.<p>• Data Sorting and Filtering: Sort data based on a column or multiple columns; use filters to view specific rows of data.<p>• Aggregation Function: Automatically summarize statistics for each column, providing instant calculations like sum, average, count, max, and min for streamlined data analysis.<p>• Data Formatting: formatting numbers, dates, etc.<p>• Grouping: Organize rows into collapsible groups based on column values for easier data analysis and navigation.<p>• Import/Export Capabilities: Import and export data from other formats, e.g., .csv, .xlsx.<p>Multiple Views<p>Visualize and interact with data in various ways best suited for their specific tasks.<p>• Grid View: The default view of the table, which displays data in a spreadsheet-like format.<p>• Form View: Input data in a form format, which is useful for collecting data.<p>• Coming soon: Kanban View, Calendar View, Gallery View, Gantt View, Timeline View.<p>Super Fast Amazing response speed and data capacity<p>• Millions of data are easily processed, and there is no pressure to filter and sort<p>• Automatic database indexing for maximum speed<p>• Supports batch data operations at one time<p>Full-featured SQL Support Seamless integration with the software you are familiar with<p>• BI tools like Metabase PowerBi...<p>• No-code tools like Appsmith...<p>• Direct retrieve data with native SQL<p>Privacy-First<p>• Bring your own database (coming soon)<p>Real-time collaboration • No need to refresh the page, data is updated in real-time

Show HN: LlamaGym – fine-tune LLM agents with online reinforcement learning

Show HN: Timelock.dev – Send a secret into the future using timelock encryption

This is simply a web interface built on top of the timelock encryption system posted by Cloudflare last week. <a href="https://blog.cloudflare.com/harnessing-office-chaos" rel="nofollow">https://blog.cloudflare.com/harnessing-office-chaos</a>

Show HN: Timelock.dev – Send a secret into the future using timelock encryption

This is simply a web interface built on top of the timelock encryption system posted by Cloudflare last week. <a href="https://blog.cloudflare.com/harnessing-office-chaos" rel="nofollow">https://blog.cloudflare.com/harnessing-office-chaos</a>

Show HN: Wallstreetlocal – View investments from America's biggest companies

Hello Hacker News! My name is Anonyo, and I am a seventeen-year-old from Southeast Michigan. This is wallstreetlocal, my passion project for the last year (and a half). I've posted this before, but I've finally open-sourced this entire project, so I thought I'd post it again.<p>Heres the short pitch.<p>The Securities and Exchange Commission (SEC) keeps record of every company in the United States. Companies whose holdings surpass $100 million though, are required to file a special type of form: the 13F form. This form, filed quarterly, discloses the filer's holdings, providing transparency into their investment activities and allowing the public and other market participants to monitor them.<p>The problem though, is that these holdings are often cumbersome to access, and valuable analysis is often hidden behind a paywall. Through wallstreetlocal, the SEC's 13F filers become more accessible and open.<p>By exploring the website (and the code), you can see the resources I used, check out some notable money managers I listed, and download any data that suits you. All for free. (Note, the mobile site likely needs work.)<p>I made this project to better democratize SEC filings, and also to get some experience on my hands. I love computers, and one day hope to get involved with startups. In the comments, I'd appreciate any and all advice, as well as feedback on how to improve the site.

Show HN: Hatchet – Open-source distributed task queue

Hello HN, we're Gabe and Alexander from Hatchet (<a href="https://hatchet.run">https://hatchet.run</a>), we're working on an open-source, distributed task queue. It's an alternative to tools like Celery for Python and BullMQ for Node.js, primarily focused on reliability and observability. It uses Postgres for the underlying queue.<p>Why build another managed queue? We wanted to build something with the benefits of full transactional enqueueing - particularly for dependent, DAG-style execution - and felt strongly that Postgres solves for 99.9% of queueing use-cases better than most alternatives (Celery uses Redis or RabbitMQ as a broker, BullMQ uses Redis). Since the introduction of SKIP LOCKED and the milestones of recent PG releases (like active-active replication), it's becoming more feasible to horizontally scale Postgres across multiple regions and vertically scale to 10k TPS or more. Many queues (like BullMQ) are built on Redis and data loss can occur when suffering OOM if you're not careful, and using PG helps avoid an entire class of problems.<p>We also wanted something that was significantly easier to use and debug for application developers. A lot of times the burden of building task observability falls on the infra/platform team (for example, asking the infra team to build a Grafana view for their tasks based on exported prom metrics). We're building this type of observability directly into Hatchet.<p>What do we mean by "distributed"? You can run workers (the instances which run tasks) across multiple VMs, clusters and regions - they are remotely invoked via a long-lived gRPC connection with the Hatchet queue. We've attempted to optimize our latency to get our task start times down to 25-50ms and much more optimization is on the roadmap.<p>We also support a number of extra features that you'd expect, like retries, timeouts, cron schedules, dependent tasks. A few things we're currently working on - we use RabbitMQ (confusing, yes) for pub/sub between engine components and would prefer to just use Postgres, but didn't want to spend additional time on the exchange logic until we built a stable underlying queue. We are also considering the use of NATS for engine-engine and engine-worker connections.<p>We'd greatly appreciate any feedback you have and hope you get the chance to try out Hatchet.

Show HN: Flyde – an open-source visual programming language

Hi HN! I’m Gabriel, and I’m happy to share a project I’ve been working on for the last few years: Flyde, an open-source visual programming language. Check out the interactive examples and online playground on the website: <a href="https://www.flyde.dev" rel="nofollow">https://www.flyde.dev</a>.<p>In my last role as an engineering manager for a B2B-oriented product, I authored and reviewed many diagrams for backend applications, mostly for integrations between 2 third-party services. Some of these diagrams were elaborate enough that I started dreaming of a way to simply run a diagram as is; I imagined a “run” button on the top-right corner of the screen that would execute the diagram without the need to translate it into code.<p>That led me down a rabbit hole of exploration and experimentation, from tools like Zapier, Pipedream and Make, which are great for automating “backoffice” stuff, and up to NodeRED, NoFlo.js and the great work of J. Paul Morisson on Flow-Based Programming. I failed to find a tool that would answer my needs - a tool that balances a new level of abstraction, manages to stay powerful and flexible, and most importantly, integrates with the existing ecosystem, and doesn’t replace it. I built Flyde as an attempt to answer that need.<p>Flyde is designed to complement and enhance traditional textual coding, not to replace it. It includes a VSCode extension, it seamlessly integrates with existing TypeScript/JavaScript code and can run on Node.js and in the browser.<p>I believe that as we delegate more coding tasks to AI, we’ll assume the role of an architect rather than a programmer. This shift will require tools that focus more on orchestration and high-level troubleshooting and less on low-level functionality.<p>I’d love to hear your thoughts and feedback on Flyde’s direction!

Show HN: Fructose – LLM calls as strongly typed functions

Hi HN! Erik here from Banana (formerly the serverless GPU platform), excited to show you what we’ve been working on next:<p>Fructose<p>Fructose is a python package to call LLMs as strongly typed functions. It uses function type signatures to guide the generation and guarantee a correctly typed output, in whatever basic/complex python datatype requested.<p>By guaranteeing output structure, we believe this will enable more complex applications to be built, interweaving code with LLMs with code. For now, we’ve shipped Fructose as a client-only library simply calling gpt-4 (by default) with json mode, pretty simple and not unlike other packages such as marvin and instructor, but we’re also working on our own lightweight formatting model that we’ll host and/or distribute to the client, to help reduce token burn and increase accuracy.<p>We figure, no time like the present to show y’all what we’re working on! Questions, compliments, and roasts welcomed.

Show HN: dockerc – Docker image to static executable "compiler"

Show HN: Astro App

I really like Stellarium and SkySafari but I felt like these are primarily geared towards exploring the sky but not so much "here are the long list of things I want to see, when can I see them tonight and where". There's also not really a great option I've found that combines sky object planning + location weather details while still being free so I built this. The UI's heavily inspired by NINAs sky atlas + Robinhood.<p>Right now you can:<p>View the altitude chart of objects and 3D view<p>Create lists of objects of interest<p>View the annual max/min daily altitude of an object to find the best time of year to view<p>See live clouds from GOES satellite view + weekly night-centric forecast

Show HN: 3 years and 1M users later, I just open-sourced my "Internet OS"

Show HN: Predictive text using only 13kb of JavaScript. no LLM

Show HN: Struct – A Feed-Centric Chat Platform

Hi HN! I’m Jason, a product designer at Struct Chat.<p>At Struct, we're frustrated by the clutter, distractions, and inefficiencies plaguing existing chat platforms like Slack and Discord.<p>We've built a radical new chat platform that leverages threads, feeds, and AI to help alleviate these problems, and give you back more time in your day.<p>Struct uses a thread-first approach to keep conversations on-topic. It applies AI-generated titles and summaries to help you decide what's worth your attention. Threads are then organized in a real-time feed, keeping all your conversations up-to-date and at the ready, eliminating the need for channel hopping.<p>Comprehensive search tools make finding things a breeze, and Strucbot, our AI assistant can answer questions based on what it’s learned from prior threads. It can even proactively respond when it notices repeat requests, providing quick answers so you don’t have to. Structbot is fully GPT-4 enabled, so you can riff with Chat GPT and your peers (generate code, ask questions, all the good stuff) without ever switching apps.<p>Struct is available on Linux, Windows, Mac, and even works as a Slack interface. Give us a try and let us know what you think.

< 1 2 3 ... 20 21 22 23 24 ... 122 123 124 >