Welcome to the third and final publication of our articles dedicated to the development of RAG applications based on LLM models. In this final article, we will see, based on our small example project, how we can find the most appropriate context for the question we want to send to our LLM model and for this we will make use of the vector search functionality included in IRIS.
My main goal of this article was to prove the use of InterSystems IRIS for Health for REST FHIR interoperability between multiple applications. In this use case, some initiating application makes a REST call to IRIS for Health (which is merely a passthrough for REST calls) to retrieve FHIR data from an Oracle Health R4 FHIR repository. Ideally, it simplifies the syntax for calling the Oracle Health APIs.
Recently, I come up an idea in my mind that how can I put my playlist on IRIS.🧐
At the same time, I was told to pay for my Spotify subscription💸💸... ooo.. how about to get some data from the Spotify API... so I started to do study about it.
I'm trying to convert date - 2023-09-28T20:35:41Z to BST/GMT format. I tried with $ZDT($ZDTH("2023-09-28T20:35:41Z",-2),8,1) but it's giving the output as '19700101 01:33:43' and looks link the date format in $ZDTH specified is wrong. Any inputs or solution would be appreciated.
Watch this video to find out how these digital health startups are improving workflows for disease management, remote patient monitoring, and patient engagement:
https://www.youtube.com/embed/-_lZ1CGqB9I [This is an embedded link, but you cannot view embedded content directly on the site because you have declined the cookies necessary to access it. To view embedded content, you would need to accept all cookies in your Cookies Settings]
I´m working with Cache for many years, currently i have running version cache-2018.1.7.721.0-win_x64.exe
However we are facing some compatibility/requirements challenges, in that way i´m hereby request your support to provide next version/build of Cache 2018.1 (e.g cache-2018.1.9.xxx.x-win_x64.exe )
Nevertheless , I already downloaded IRIS for testing purpose if can also provide trial/limited licence for detail testing will be apreciated.
When developing productions there are many places that we input endpoint information (IP/port) especially when there are dozens of operations going to the same environment. What later happens is that the IP changes and we have to then go into each operation and update the IP.
Accessing an Azure cloud storage to upload/download blobs is quite easy using the designated %Net.Cloud.Storage.Client class API methods, or using the EnsLib.CloudStorage.* inbound/outbound adaptors.
Note that you'll need to have the %JavaServer External Language Server up and running to use the cloud storage API or adaptors, since they both use the PEX framework using the Java Server.
When we create a FHIR repository in IRIS, we have an endpoint to access information, create new resources, etc. But there are some resources in FHIR that probably we wont have in our repository, for example, Binary resource (this resource returns a document, like PDF for example).
I have created an example that when a Binary resource is requested, FHIR endpoint returns a response, like it exists in the repository.
The world of Generative AI has been pretty inescapable for a while, commercial models running on paid Cloud instances are everywhere. With your data stored securely on-prem in IRIS, it might seem daunting to start getting the benefit of experimentation with Large Language Models without having to navigate a minefield of Governance and rapidly evolving API documentation. If only there was a way to bring an LLM to IRIS, preferably in a very small code footprint....
I have a question about using OnInit() within a Ens.BusinessOperation.
When you include OnInit(), does OnInit() only execute when you start a Business Operation? Or does it execute OnInit () every time you send a REST request to the operation. I am trying to pinpoint when the best time is to execute the POST command to get the Token
I am needing to get a Bearer Token from a REST POST call and return the Authorization key prior to making the rest of the REST calls to pull down data.
Since version 1.93 of VS Code, the “import and compile” process has been very slow (more than an hour). Have you encountered this issue and do you know how to resolve it?
I'm trying to learn M programing for an Epic db class prerequisite. They said to download the IRIS Studio software to do the testing. I'm having a very difficult time finding information the language. I'm trying to run some examples that Epic has provided (like in below) but the compiler complains that it isn't valid. Of course, it doesn't tell you why it isn't valid.
in the InterSystems management portal, I am in the lookup table viewer (interoperability>lookup Tables, I would like to use import function to convert our existing table from other system to intersystems lookup table.
so I converted our existing one like below and saved it as xxx.xml file locally. when I try to import, it says it is not a valid export file.
my question is " what's the valid format for a file to be valid to use to import for lookup table?"
docker login -u=<username> -p=<password> containers.intersystems.com WARNING! Using --password via the CLI is insecure. Use --password-stdin. Login Succeeded
$ docker pull containers.intersystems.com/intersystems/irishealth:2024.1 Error response from daemon: unauthorized: The client does not have permission for manifest: Download request for repo:path 'docker-customer-remote-cache:intersystems/irishealth/2024.1/list.manifest.json' is forbidden for user: 'anonymous'.
I don't have a current project where I am needing to create a new operation that uses SOAP at the moment but as someone who still has some SOAP connections in her Production, I was curious if VSCode had an equivalent to the SOAP Wizard that helped create all of the SOAP classes needed based on the provided WSDL? Does anyone know? Or has this become a manual process?
Note: this was originally posted on June 5, 2024 but presented as being posted on May 9, 2024 so this re-post fixes the date.
Recent updates to the Intersystems Language Server introduce many significant enhancements aimed at improving developer experience and productivity. I'll talk about some of the key ones here, while the complete list, including numerous bug fixes, can be found in the Language Server's CHANGELOG.
Hi, I'm working on a large extraction from a database and need to parallelize the processes during the extraction.
Here's what happens:
The user starts the extraction from.
Several jobs are started using the Job <Method> instruction.
At the end, the user expects to find a document containing the results of all the extractions. What I'd like to do is start a new job that checks whether the previously started jobs have finished or are still working and consequently produce the document.
This is a detailed, candid walkthrough of the IRIS AI Studio platform. I speak out loud on my thoughts while trying different examples, some of which fail to deliver expected results - which I believe is a need for such a platform to explore different models, configurations and limitations. This will be helpful if you're interested in how to build 'Chat with PDF' or data recommendation systems using IRIS DB and LLM models.
https://www.youtube.com/embed/bcu1gt0BDhY [This is an embedded link, but you cannot view embedded content directly on the site because you have declined the cookies necessary to access it. To view embedded content, you would need to accept all cookies in your Cookies Settings]
Year
Open Exchange app
.png)
.png)
.png)