I think you are in the wrong channel @mubarakkhan - #C021JANJ6TY might be better

Okay

You welcome

Hello Everyone! I'm a new villain who just joined this channel. I am a senior artificial intelligence engineer with 12 years of rich experience in artificial intelligence. In particular, computer vision, natural language processing and backend development are my main fields of development, and I have development experience and leadership experience in various companies, and I have very strong project development ability. I develop very fast and provide clean quality code. I want to exchange experiences with you and give you the help you need. Thank you.

A nice update from InfoWorld on the state of locally runnable LLMs. https://www.infoworld.com/article/3705035/5-easy-ways-to-run-an-llm-locally.html

Hello Everyone! Introducing a personal project I recently developed. https://7d14d25e6858e8e70f.gradio.live Music Generation Project.

Can the airbyte no code connector use the oauth2.0 auth method?

anyone using the pinecone destination connector so far and can share how they are making use of it?

Hi Team, if we are using Airbyte to sync data from Snowflake to Kafka, do we need to provide larger snowflake warehouse for extracting large volume of data from snowflake? Or the snowflake warehouse size does not matter for syncing data(large or small volume) from snowflake via Airbyte

EDIT – RESOLVED: This was solved by having

credentials_title

be exactly

"Private App Credentials"

. hi all! I’m currently following the LangChain guide for the

AirbyteHubspotLoader

but running into an issue when trying to connect to my Hubspot private app. This is my code:

config = {
  "start_date": "2023-08-20T00:00:00Z",
  "credentials": {
    "credentials_title": "<my-title>",
    "access_token": "<redacted>"
  }
}

loader = AirbyteHubspotLoader(config=config, stream_name="contacts")

However, I get this error when running it and am at a loss:

File "/opt/homebrew/anaconda3/lib/python3.10/site-packages/airbyte_cdk/sources/utils/schema_helpers.py", line 175, in check_config_against_spec_or_exit
    raise AirbyteTracedException(
airbyte_cdk.utils.traced_exception.AirbyteTracedException: 'OAuth Credentials' was expected

Seems like the loader is expecting OAuth Credentials, but I just have a Private App. Is there anything I’m doing wrong? (also let me know if this isn’t the best place to ask this question)

Anybody experimenting with RAG apps? Interesting discussion of various approaches to chunking and strategies for improving results. https://www.ninetack.io/post/improving-rag-quality-by-summarization

hey hey , quick question: does Airbyte support writing to GSheet from trino query?

Hi guys, I am facing this issue almost everyday.

Does anyone have any idea?

This is deployed on ec2

Hey guys , so I have quick question …. When clonking the air byte project I notice there are some files missing like the docker-compose , .env does running the .sh file install them or ??

Everyone from Airbyte Team was removed from this channel because @[DEPRECATED] Marcos Marx archived the channel.

archived the channel

set the channel topic: Channel to share and discuss AI 🤖

Really excited to see what people bring to this channel. Holler if anyone has cool AI projects to share!

Glide has joined this channel by invitation from Airbyte.

During the weekend I shows how to build a AI Chat app using n8n+supabase. I found it interesting that Supabase supports the Postgres Vector Store, and their API is so simple to use.

An interesting read on the ai landscape, " lab companies Vs wrappers " https://yellow-apartment-148.notion.site/The-Zero-Day-Flaw-in-AI-Companies-b164d5eb97324a6d80252ee6b726d3eb

Speaking of o1, the system card released with it is pretty interesting. Specifically the whole faking alignment when there is high oversight. Will be interesting as we learn more about how AI systems determine appropriate ways to act when oversight isn't perfect

I’ve felt like this is going to be a trend many others will follow, but then the pendulum will swing quickly back to using SaaS providers due to implementation expertise. That or we’ll find a growing category of jobs for outlining the org, the software it needs, and then the support services. I used to run an HR SaaS company and, crazy enough, most of our time was actually on designing processes/trainings for adoption, rather than development.

Unsure if people have played with Google's NotebookLM. I put some articles through it to test the audio feature/make a podcast. Thought it was pretty cool, consider checking it out

Everyone from Glide was removed from this channel because @[DEPRECATED] Marcos Marx archived the channel.

archived the channel