As an absolute terrible front-end developer, I’ve completely fallen in love with Streamlit – it makes building apps so simple, without the hassle of handcrafting a UI from scratch 🎨.

I’m currently building an app that transcribes, summarises and translates audio using the AI capabilities of OCI (more on this in a future post 🗓️).

One of the things I wanted to do in the app is record audio, thankfully Streamlit has st.audio_input, which can record audio from a users’ microphone, however I needed a way to save the recorded audio to the server running the Streamlit app so that I could work some AI-magic on it 🪄.

It turns out that this is super-simple, the code below is for a page that has the st.audio_input widget, when a recording has been created it is saved with the filename recorded_audio.wav.

import streamlit as st

st.sidebar.title("Audio Recording App")
st.title("Record Your Audio")
st.write("Press the button to start recording and then stop when you're done.")
audio = st.audio_input("Record your audio")

if audio:
    with open("recorded_audio.wav", "wb") as f:
        f.write(audio.getbuffer())
        st.write("Audio recorded and saved successfully!")

Here’s the page in all it’s glory:

Within the directory that the Streamlit app is run from you can see that a WAV audio file has been saved:

Happy days 😎.

I’d deployed a shiny new Function to OCI, I went to test this using fn invoke and it failed with the following error:

Error invoking function. status: 502 message: Failed to pull function image

The reason for this error is that I didn’t have a Service Gateway provisioned within the VCN that hosted the function app – but what is a Service Gateway you may ask???

A service gateway lets a Virtual Cloud Network (VCN) privately access specific Oracle services without exposing the data to the public internet. No internet gateway or NAT gateway is required to reach those specific services. The resources in the VCN can be in a private subnet and use only private IP addresses. The traffic from the VCN to the Oracle service travels over the Oracle network fabric and never traverses the internet.

To fix this I created a Service Gateway and attached this the subnet where my function app had been deployed, this provided the function app with access to the Container Registry to pull the image that contained the function.

In my case I’d deployed the function app to a private subnet that had no connectivity to the outside world – even to OCI services 🔐.

I then needed to add a rule to route traffic through the Service Gateway.

Once I’d done this, the function worked 💪.

I was going through the process of configuring OCI to send audit logs to Microsoft Sentinel using the following walkthrough – https://docs.oracle.com/en/learn/stream-oci-logs-to-azure-sentinel/

When I got the section to configure Azure (Task 5), I ran into an issue – it wasn’t clear exactly what I needed to populate the App Insights Workspace Resource ID setting with as it’s not covered within the documentation 🤔.

This setting can be obtained from the Log Analytics workspace in Azure that is created in Task 4.

Go to Settings > Properties:

Copy the Resource ID and paste this into the App Insights Workspace Resource ID setting.

Once I’d done this, I was able to successfully configure the integration and now have lots of lovely OCI audit logs within Microsoft Sentinel.

Select AI is a fantastic capability included in Oracle Autonomous Database, which enables a database to be queried using natural language rather than SQL commands, which is a godsend for somebody like me who struggles with all of that select * from malarkey!

I was going through one of my Select AI demos in preparation for a customer meeting and all my queries were failing with the error below – typical, just before I needed to demo it 😫.

ORA-20400: Request failed with status HTTP 400 – https://inference.generativeai.us-chicago-1.oci.oraclecloud.com/20231130/actions/chat Error response – { “code”: “400”, “message”: “Entity with key cohere.command-r-plus not found” } ORA-06512: at “C##CLOUD$SERVICE.DBMS_CLOUD”, line 2100 ORA-06512: at “C##CLOUD$SERVICE.DBMS_CLOUD_AI”, line 10811 ORA-06512: at line 1 https://docs.oracle.com/error-help/db/ora-20400/

The reason for this error is that the Select AI profile that I was using was configured to use the cohere.command–r-plus LLM from the OCI Generative AI Service, this LLM has been deprecated as per the documentation, therefore no longer works:

To fix this issue I needed to update (well delete and re-create!) the Select AI profile to use the newer variant of this LLM, which is cohere.command-r-plus-08-2024.

Deleting a profile – replace ProfileName with the name of the profile to delete 🗑️

BEGIN
     DBMS_CLOUD_AI.DROP_PROFILE(profile_name => 'ProfileName');
END;

Re-adding the profile with the new LLM – Example ➕

BEGIN                                                                        
  DBMS_CLOUD_AI.CREATE_PROFILE(                                              
      profile_name => 'OCI_COHERE_COMMAND_R_PLUS',
      attributes   => '{"provider": "oci",
                        "credential_name": "GENAI_CRED",
                        "object_list": [{"owner": "ADMIN"}],
                        "model": "cohere.command-r-plus-08-2024"
                       }');
END;

Buried within the December 2024 release notes for the OCI Generative AI Agents service is this little gem 💎:

This now enables you to do some prompt engineering to influence the response produced by an agent, this is useful if you need to tailor the length, style and tone of the response from the agent. For example you may need the response to include a maximum of 3 bullet points.

To provide additional instructions to an agent response, navigate to the agent and select Edit. Below you will see a field named Instructions for RAG generation, within this add the additional instructions. In the example below, I have simply asked it to crerate a short summary using a maximum of 3 bullet points.

Here is a before/after comparison of the response from the agent, this is using an agent I built that is trained on UK immigration policy data.

Before

After

There’s some other interesting features in the December release, including more detailed citations and the ability to override Object Storage citation links through custom Object Storage metadata.