Jump to content

Search the Community

Showing results for tags 'rest'.

  • Search By Tags

    Type tags separated by commas.
  • Search By Author

Content Type


Forums

  • Caspio Bridge
    • Caspio Apps for Ukraine
    • General Questions
    • Caspio JavaScript Solutions
    • Tables, Views and Relationships
    • Import/Export and DataHub
    • DataPages
    • Deployment
    • Security, Authentications, Roles, SAML
    • Styles and Localizations
    • Parameters
    • API and Integration
    • Calculations and aggregations
    • User JavaScript and CSS Discussions

Find results in...

Find results that contain...


Date Created

  • Start

    End


Last Updated

  • Start

    End


Filter by number of...

Joined

  • Start

    End


Group


MSN


Website URL


ICQ


Yahoo


Skype


Location


Interests

Found 10 results

  1. I've noticed that the topic of using AI in Caspio apps hasn't gotten much attention on this forum. This is surprising, given the big advancements in AI over the past few years. Recently we discussed this topic during the Partners’ Office Hours with @nikcaspio and decided to test the water sharing some ideas on the forum and hearing what the community thinks. I'd gladly give more examples in future posts if you find these thoughts helpful. The topic is almost endless, with so many ways to use AI. But let's start simple: WHAT IS DOCUMENT UNDERSTANDING We all know how Caspio can help create outgoing PDF documents based on database records. But what about incoming documents like invoices, receipts, delivery notes, or bills of lading? These often come as files or even paper documents. Here, we have to do the opposite of document creation. We need to take the document and turn it into data. Usually, someone has to enter this data manually into Caspio, right? Well, that was the old way Things have changed. To show what's possible now, I made a short video using our Caspio app, where I upload a JPEG file with an invoice, and the system extracts the necessary data from it and put it into Caspio: https://www.loom.com/share/d21eb53d9b5a4540a6282a56d8d47f0e Is it magic? No, it's simply artificial intelligence. And a pretty inexpensive one. I paid only a few cents to our "digital elves" for assisting me in recognising these documents in the video. The cost of human labor would most likely be much higher. TECH STACK Of course, you first need a Caspio plan with REST API and Zapier Integration. The Professional plan offers it, but you may buy it as an add-on to smaller plans. In this example, we used Google Document AI's Invoice Processor to do the work. It has some limits but is a good place to start and see quick results. And you can overcome these limits with more advanced methods. However, for this demo, it's more than enough. At first, I considered using Zapier for this demo because it was easy. But I know many people here find Zapier too expensive. So, we chose a more budget-friendly option. We used the newly introduced Caspio Webhooks and Make (Integromat) as middleware. This is just an example; you can use any technology you like. MAIN COMPONENTS: To make Document Understanding work, you need three main parts: OCR (Optical Character Recognition): This tech changes a binary file (like a photo taken with your phone) into text. Text Extractor: This tool looks for specific info (like invoice numbers or dates) and gives it back in an organized way. Middleware: This is the connector between Caspio and AI services. It takes a file from Caspio, sends it for processing, returns the data, and puts it into a Caspio table. GOOGLE DOCUMENT AI Google Document AI can do tasks #1 and #2. You can learn more about this service here and here's how to set it up. Of course, Google Document AI is not the only option, but it's an excellent place to start. Once you know its limits, you'll know how to work around them. There are many ways to do this. For example, in most of our projects, we use Google Document AI mainly for OCR. Then, we process the text with a tool from OpenAI, the creators of ChatGPT. But that's more advanced, and I don't want to make things too complicated in this article. MIDDLEWARE Google Document AI comes with a robust and well-explained API. But it needs incoming calls to have a particular format and gives back answers in its own format. If you're already using Zapier with Caspio, you might find it the easiest option, and you probably know how to use it. If so, skip to the Make Scenario section below for extra ideas. But if you're not using Zapier, stay with me. This post will focus on Caspio Webhooks as a more budget-friendly choice. Webhooks are good, but they have a limit: they're set in a standard way and don't allow much change. So, you'll need another tool to take the webhook call from Caspio and change it into a format that Document AI understands. There are many choices for this, and it's a big topic that could have its own set of detailed articles. For this demo, we'll use Make (Integromat). We find it even better than Zapier, but cheaper. They also have an excellent free plan that might be enough for smaller projects. Finally, Caspio has built a connector for Make, making it easy to set up, even if you're not an API expert. CASPIO CONFIGURATION To keep it easy, we turned on the FileStor option. This lets us get the file from Caspio using a direct link. I know this isn't the most secure way to handle files, but that's not the main point of this article, so I hope you understand. We also made a table and a few data pages in our demo app. These let you create a record and upload a file. Next, we set up a Caspio webhook that starts when a new record is added (Insert event). This webhook sends the call to Make's incoming webhook. Finally, we made a Web service profile so Make can talk to Caspio using REST API and return the results to the table. All these steps are well-covered in the Caspio manual, and there are also good video tutorials. So, I won't go into the details here. (If you need those, just ask in the comments, and I can post the links.) SCENARIO IN MAKE The process has just six steps in order: An incoming webhook that waits for calls from Caspio. A 10-second pause to give Caspio time to send the file to FileStor. A simple GET request to get the file from FileStor. Some minor changes to the data to get it ready for Document AI (details below). A call to Document AI to send the prepared data. A standard Caspio connector that puts the data back into the Caspio table. If you're new to Make or APIs and find all of this confusing, let me know if you'd like more details. I intended this post to be an outline rather than a step-by-step guide. However, I can explain things further if needed. It's not rocket science. But If you're familiar with Make or at least REST API, most of these steps should be pretty straightforward. The only part that might need some extra attention is step #4, where we set up headers and converted the binary file to base64. Here's how we set it up: If you've done everything right, you should be good to go! Not too hard, right? MORE ON THE TOPICS In this post, we've just touched the tip of the iceberg on Document Understanding. It's a big topic that could fill many more articles and take hundreds of hours to make a really deep dive. This is especially true for understanding complex documents like contracts or unstructured text documents. And the real magic happens when you move past ready-made solutions and train AI models with your own data. Learning all this can be a fun adventure, and the possibilities are endless. However, even with simple setups, AI can be a game-changer for many businesses. It can save time on dull tasks and let your team focus on more important work instead of just copying data from paper to an app. But AI isn't just about Document Understanding; it has lots more to offer. Here are a few more examples: Semantic Search: This isn't just looking up keywords; it understands the meaning of your question, irrespective of the words you choose. This can be useful for Knowledge Management and Customer Support apps. Voice-to-Text: Turn your video calls into text, summarize what was said, and add it to your Caspio app as follow-up notes. This is good for CRM, Project Management, Knowledge Management, and Recruitment apps. Classification of Incoming Requests: Automatically sort new requests and applications based on its content. This can help with CRM, Customer Support, and Recruiting apps. And there's much more. FINAL REMARKS I must say, this article ended up being longer than I first thought it would be. So, big thanks to you for sticking with me till the end! If anything was unclear or you ran into issues, please feel free to leave a comment or write me a DM if you prefer. I'll do my best to help you out. Now, I'd really like to hear your thoughts. Did I explain it well, or did it just confuse you? Does it make sense for you? Are you interested in this topic at all? Do you want to see more articles like this one? If so, what would you like me to focus on? More technical details? Real-world examples? A different writing style? I welcome your feedback and constructive criticism. Thank you!
  2. I am using Caspio Restful API to insert record. It returns a status code 201 which means it is inserted. I can also verify that it is inserted. However is it a normal practice to return a JSON array once inserted? Currently, it is sending an empty array. ' Any thoughts about this?
  3. Hello! I was wondering if anyone had any trouble when refreshing their access tokens? I successfully authenticated my users through RESTful API, but now have trouble using the POST call advised from the how to guide on the website. Any feedback is greatly appreciated!
  4. I have a client that has asked if they can create an .sde connection directly with my caspio database rather than utilizing the API. Below is the detailed request. This is the point in s question - "we'd need an accessible endpoint directly into your database, or at least have port forwarding set up." Is this possible? The SDE file creation needs the following information: database_platform - takes a few different platforms, the most notable being ORACLE, POSTGRESQL, and SQL_SERVER instance - The reachable address and port number we can use to connect from our ArcGIS server to your Database. This is the hard one - we'd need an accessible endpoint directly into your database, or at least have port forwarding set up. Username - the database username we'd use to connect This user would have read-only privileges only on the tables where we have data stored. Password - Password for above database user database - Only needed if running on a POSTGRESQL server, the name of the database
  5. How can I authenticate user passwords against a "password" type field in a Caspio view via the REST API? "Select" does not work with fields of type "Password" and I can't seem to figure out how to access the data. I'm writing a mobile application that uses the REST API to access our data via "views". One of the fields is a "password" field, and I need to be able to compare the password that the user enters into my application with the password stored in the database. I don't need to extract the password field, but I do need to know if the password entered by the user matches what's stored in the database. Any tips on how to do this, or links to documentation, would be greatly appreciated.
  6. Hello Caspio Community! Let me explaining my scenario firstly please. Currently I'm stuck on how to authenticate REST API calls in a secure manner. I have a new app stood up and ready to launch (this app is/well always be locked behind Caspio Login) . However, before I launch this app I was wondering if I could use "document.cookie()" user token to make API calls instead of the bearer token that I have been using for development. If I could use a user token instead of an API key I would feel much more secure. Thank you for taking time out of your day to read this! I look forward to reading your replies!
  7. Hi, Greeting, I am new to this forum, I am having a problem when inserting multiple records at once by REST. It works perfect while inserting single record each time. However, it takes more than 3 mins when inserting about 300 records, does anyone know how to make a bulk insert instead of looping it. I have tried the following, but it doesn't work. [{"ID":"1","Date":"11/29/2016"},{"ID":"2","Date":"11/30/2016"} ] {"ID":"1","Date":"11/29/2016"},{"ID":"2","Date":"11/30/2016"}
  8. This may be of use to people. This is pulled out of a project I'm working on. It's a PHP library that manages the low-level integration with the Caspio REST API. If I have time I'll pull more classes out of the project that might be helpful and integrate them into this library. But it demonstrates the basics of interacting with the Caspio REST API via PHP. https://github.com/moksamedia/caspio-rest-api-lib
  9. I need to download more than 1000 rows via the REST API, is there an easy way to do it? Something like "get next 1000 rows", etc.? We've written an application that downloads data to a tablet via the REST API, the user then takes the tablet to remote locations off-line. Some of the data tables have more than 1000 rows, so our users are not able to get all of their data. I'm hoping Caspio has a built in way to "get next 1000 rows" so that we don't have to alter our tables to have sequential row indexes.
  10. I created a profile. Put my client_id and client_secret into POSTMAN along with the parameter "grant_type" and the value "client_credential". All I can get from Caspio is { "error": "invalid_client" } What other information do I need to include with the request? I've read the documentation. http://howto.caspio.com/caspio-bridge-rest-api/authenticating/ grant_type=client_credentials&client_id=<Client Key>&client_secret=<Client Secret> That list of URL Parameters is exactly what I'm sending. I've gone back into Caspio Access Permissions and refreshed my client_id and client_secret and saved the changes. With the new id and secret, I'm getting exactly the same error message. Please Help
×
×
  • Create New...