OK! Let’s dig in.

Welcome @Levent Gorur! Glad you were able to jump in.

Welcome, @Levent Gorur. If you have any questions about LangStream, this is the place to ask!

Hello everyone, I hope each of you has fruitful outcomes.

Hello folks

Here you can find LangStream and try it out https://github.com/LangStream/langstream

Documentation: https://docs.langstream.ai/

Looks like I get to ask the first question. I know there's a concat compute transform, but I'm looking for a way to use a compute transform (or another if there is one) to split a string. There appears to be no split EL function and method invocations are disabled so I can't use the Java String split function. Ideas?

Here you go, new utility functions 😁 https://github.com/datastax/pulsar-transformations/pull/117

Our website and first blog post is live! Please amplify them on social media. And if you haven't already, star the GitHub repo! https://x.com/langstream_ai/status/1702034257223250423?s=20 https://www.linkedin.com/feed/update/urn:li:activity:7107801120076414976

<!here> 🔔 we just released LangStream 0.0.17 with new cool features: • New

chat

gateway which simplifies the usage of the gateways • Introduced support for Azure Blob Storage as code storage • New tool for testing the application locally without kubernetes Checkout the full release notes: https://github.com/LangStream/langstream/releases/tag/v0.0.17 Subscribe to the project releases by clicking Watch -> Releases: https://github.com/LangStream/langstream and please add your ⭐ to spread the project

I hadn’t tried it before, but actually the LLM is so smart that I can issue questions in Italian 🇮🇹 This is amazing ! I have just deployed the WebCrawler demo application and followed the instructions. https://github.com/LangStream/langstream/tree/main/examples/applications/webcrawler-source

You:

> What is Lang Stream ?

....Server:

LangStream is an open source project that combines the power of large language models. with the agility of streaming processing to create powerful processing applications. LangStream applications are.. designed to be able to watch a message topic and process data through multiple stages to output useful generative AI results. This means that LangStream can be used to build resilient and production-ready chatbots,.. recommendation systems and other AI applications. The LangStream project involves multiple systems and components, and it takes monitoring to ensure their health and quick mitigation.

You:

> Che cosa è LangStream ?

..Server:

LangStream è un progetto open source che unisce l’intelligenza dei grandi modelli di linguaggio con l’agilità del processing.. per creare potenti applicazioni di elaborazione di dati in tempo reale. In pratica, LangStream permette di creare applicazioni in grado di elaborare flussi di dati.... continuamente aggiornati mediante una combinazione di diversi sistemi, come ad esempio chatbot, agenti di elaborazione testi e molto altro. Il progetto LangStream util.izza tecnologie avanzate per l’elaborazione di testi, il rilevamento del linguaggio e la manipolazione di dati in tempo reale in modo efficiente e scalabile.

hey folks, I'm trying to get the sample-app depoyed, but running into issues

sample-app langstream apps get sample-app
The operation couldn't be completed. Unable to locate a Java Runtime.
Please visit <http://www.java.com> for information on installing Java.

ID               STREAMING        COMPUTE          STATUS           EXECUTORS        REPLICAS
sample-app       kafka            kubernetes       ERROR_DEPLOYING  0/0

this is the error in the logs,

Couldn't resolve server my-cluster-kafka-bootstrap.kafka:9092 from bootstrap.servers as DNS resolution failed for my-cluster-kafka-bootstrap.kafka

sample-app kubectl get services -n kafka
NAME                          TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)                      AGE
my-cluster-zookeeper-client   ClusterIP   10.100.160.164   <none>        2181/TCP                     21m
my-cluster-zookeeper-nodes    ClusterIP   None             <none>        2181/TCP,2888/TCP,3888/TCP   21m

i didn't find

my-cluster-kafka-bootstrap.kafka

in kafka namesapce..

ok starting a new thread .. sample-app is now deployed, but http error while consuming output I can see the port-forwarding going thro for 8090 and 8091 .. any tips to troubleshoot this one

langstream gateway consume sample-app consume-output -p sessionId="$session"
The operation couldn't be completed. Unable to locate a Java Runtime.
Please visit <http://www.java.com> for information on installing Java.

The HTTP request to initiate the WebSocket connection to [<ws://localhost:8091/v1/consume/default/sample-app/consume-output?param:sessionId=20192F2C-EC50-430C-BD97-1263533126E9>] failed

i'm trying to run locally, with

LangStream CLI 0.0.19 (f0d59d56)

its the sample app from here (just deleted the

instance.yaml

- since the run complained not to include it)

langstream docker run myapp -app . -i <https://raw.githubusercontent.com/LangStream/langstream/main/examples/instances/kafka-docker.yaml> -s /tmp/secrets.yaml

it looked fine, but unable to send message

➜  ~ session="$(uuidgen)"
➜  ~ langstream gateway produce myapp produce-input -p sessionId="$session" -v "Barack Obama"
The operation couldn't be completed. Unable to locate a Java Runtime.
Please visit <http://www.java.com> for information on installing Java.

No gateway defined for application myapp

do i need any other config

i wanted to just use openai as the provider, but ran into this issue

com.azure.core.exception.ResourceNotFoundException: Status code 404, "{
  "error": {
    "message": "Invalid URL (POST /openai/deployments/gpt-35-turbo/chat/completions)",
    "type": "invalid_request_error",
    "param": null,
    "code": null
  }
}
"
	at com.azure.core.implementation.http.rest.RestProxyBase.instantiateUnexpectedException(RestProxyBase.java:347)
	at com.azure.core.implementation.http.rest.SyncRestProxy.ensureExpectedStatus(SyncRestProxy.java:130)
	at com.azure.core.implementation.http.rest.SyncRestProxy.handleRestReturnType(SyncRestProxy.java:213)
	at com.azure.core.implementation.http.rest.SyncRestProxy.invoke(SyncRestProxy.java:81)
	at com.azure.core.implementation.http.rest.RestProxyBase.invoke(RestProxyBase.java:109)
	at com.azure.core.http.rest.RestProxy.invoke(RestProxy.java:91)
	at jdk.proxy3/jdk.proxy3.$Proxy185.getChatCompletionsSync(Unknown Source)
	at com.azure.ai.openai.implementation.OpenAIClientImpl.getChatCompletionsWithResponse(OpenAIClientImpl.java:971)

this is my configuration.yaml

configuration:
  resources:
    - type: open-ai-configuration
      name: OpenAI Azure configuration
      configuration:
        url: "{{{ secrets.open-ai.url }}}"
        access-key: "{{{ secrets.open-ai.access-key }}}"
        provider: openai
  dependencies: []

and secrets.yaml

secrets:
  - name: open-ai
    id: open-ai
    data:
      url: <https://api.openai.com>
      access-key: xxxx

i understood that url was not really required for openai provider

/openai/deployments/gpt-35-turbo/chat/completions

this doesn't look like a openai path

thanks @Chris Bartholomew @Enrico Olivelli @Nicolò Boschi for your help! got the sample-app working!! some notes on my run https://docs.google.com/document/d/13SmGY3c75uPDgB_YkUAc-YdH-kwL4pH9ReYJ_JXFRKg/edit

<!here> 🔔 we just released LangStream 0.0.20 with new cool features: • New deletion mode for assets and topics - automatically cleanup the environment after your application has been deleted • Web crawler agent has been improved and it can now works better with sitemaps, robots.txt and other nuances • Core improvements like CLI chat output for streaming messages, usage of the OpenAI client Checkout the full release notes: https://github.com/LangStream/langstream/releases/tag/v0.0.20 Subscribe to the project releases by clicking Watch -> Releases: https://github.com/LangStream/langstream and please add your ⭐ to spread the project

Hey all. Are there plans for an agent that makes calls to restful APIs, etc over http?

Today we have committed the initial support for Milvus.io 🥳 https://github.com/LangStream/langstream/pull/470