In an earlier blog post, we looked at how you can manage conversations using threads with the Microsoft Agent Framework. In this post, we look at how you can extend an AI agents intelligence using fu...

In my earlier blog post, we introduced the Microsoft Agent Framework. We saw how the new open-source SDK from Microsoft blends the best features from Semantic kernel, Auto Gen and the Process Framework. In this blog post we take an initial first look at the Microsoft agent framework in terms of the components that form this new open-source SDK. Specifically, we cover the following: What is an AI agent Available agent types When to use an AI agent When not to use an AI agent Conversations and threads Agent Function tools Agents as function tools Memory and memory types Middleware Background processing Observability Workflows   This blog post is about the main concepts, components and patterns that are core to shipping AI agents using the Agent Framework. Based on prior project experience, I find some of these components and tooling to be similar to others I've used in the past.  For example, ChatHistory in Bot Framework and Semantic Kernel. Thes

Innovation in recent years in the conversational and agent AI space has been wild. Microsoft have recently shipped a new open-source engine that brings together the best bits from Semantic Kernel, AutoGen and the Process Framework. In this blog post, I pull together some of the recent announcements, what it means for you, and weave in some of my own thoughts. We cover: Microsoft Agent Framework Why Agents Need a New Foundation The Paradox of Choice What this means for Semantic Kernel, Process Framework and AutoGen What This Means for Your Current Project How long will Semantic Kernel Be Support for Microsoft’s Official Position Azure AI Foundry Integration Signals to Pay Attention To The AI Timeline and Looking Ahead   An Example of the Agent Framework in Action is also included.  Let’s dig in. ~ Why Agents Need a New Foundation In these last 18 months, there have been many disconnected SDKs, platforms and approaches when it comes to shipping AI age

I am partnering with Pluralsight again to bring you another new course soon. The course is called : Vector Databases & Embeddings for Developers. This course will come in at roughly 50 minutes.   Key learning objectives will centre around: Fundamentals of embeddings and vector databases. How vector databases serve as external memory for LLMs. Using pre-trained models and APIs to create embeddings   The course will be split over 3 modules. A combination of slides, visuals, code and demo applications will rapidly accelerate your understanding of vector embeddings and databases. Code examples and a sample application will help crystalise your understanding of the theory. The course will give you practical knowledge and show you how to leverage embeddings to create intelligent AI agents. Stay tuned for further information. ~ Other Courses In the meantime, you can checkout my 60 minute course, Developing an Artificial Intelligence Strategy for Your Organ

In part 5 of this mini-series, we bring together all main components to form the end-to-end solution for the AI home security system. We also refine and refactor some of the original Python code. To recap, this home security system uses motion detection, image capture, and AI-powered person recognition to distinguish between known persons (family, friends) and potential intruders. It sends alerts when unknown persons are detected. ~ System Components It’s worth a recap of the hardware and software components that form the solution. Hardware The hardware consists of: Raspberry Pi 4 Camera module PIR motion sensor connected to GPIO Bluetooth speaker for local alerts   An old laptop is also used to host a .NET API. Software Some of the files have been renamed to ease readability, the following software components are used in the solution: audioplayer.py - audio alert functions botmessage.py - Telegram messaging functions ClipServer.py - Python server running C

This is the fourth instalment of a miniseries where you see how to build an end-to-end AI home security and cross platform system. To recap, the main requirements for this system are: Detect motion and capture photo Create message with photo attached Send message to defined Telegram bot Detect who is in the photo, if the person is me, then do not invoke the Telegram bot with a message and image   I recommend reading the following before continuing: Part 1 – Introduction, Hardware, Motion Detection and Telegram Bot Part 2 – Creating a Local CLIP Server for Image Vectorisation Part 3 – Create a Local AI Custom Vision API for Training and Matching Images   In part 4, we extend the existing Telegram bot. It will be extended to allow the management of AI image functionality with the API from part 3, from a cell phone. This means you don’t need to be at a computer to train the AI with new data or to test the image recognition capability. ~ The following detai

This is the third instalment of a miniseries where you will learn how to build an end-to-end AI home security and cross platform system. I recommend reading part 1 and part 2 before reading on. To recap, the main requirements for this system are: Detect motion and capture photo Create message with photo attached Send message to defined Telegram bot Detect who is in the photo, if the person is me, then do not invoke the Telegram bot with a message and image   In part 2, we created a locally running CLIP server.  This provided an API endpoint that could accept a single image. The input image was then converted to vectors.  When in vector format, the image data is much easier to run comparisons against. ~ .NET API Capabilities In this blog post, we implement a .NET API.  The .NET API will leverage the core CLIP functionality but will expose 2 endpoints: /api/image/train /api/image/match   The /train endpoint will let us train the system with

This is the second instalment of a miniseries where I am building an end-to-end AI home security and cross platform system. I recommend reading Part 1 before reading on.  To recap, the main requirements for this system are:   Detect motion Capture photo Create message with photo attached Send message to defined Telegram bot Detect who is in the photo, if the person is me, then do not invoke the Telegram bot with a message and image   In this blog post, we implement core functionality that lets us convert images to vectors using OpenAIs free CLIP SDK and ViT-B-32 model. These can then be used to perform real-time similarity checks against new incoming data. ~ What is CLIP CLIP (Contrastive Language–Image Pre-training) is an AI model developed by OpenAI that can understand the relationship between images and text. CLIP makes it easy for you to encode images into high-dimensional vectors (embeddings). With images in vector format, you can then perform:

🚀 Ship SaaS Fast — .NET SaaS BoilerplateLaunch your SaaS in weeks, not months.Skip 3-6 months of boilerplate development and start with a battle-tested, production-ready platform. Focus on your unique...

In an earlier blog post I detailed how to create basic home security system using Raspberry Pi Zero 2.   In this blog post, I show how to create an AI home security system using .NET, Python, Semantic Kernel a Telegram Bot, Raspberry Pi 4 and Open AI.   This is part 1 of a 2-part series.   In part 1, we learn about: Required hardware and software Main requirements Installing and testing the camera module Installing and testing the PIR sensor Creating Python code to detect movement Creating Python code to capture an image Creating Python code sends a notification and image to a telegram bot   It’s a lot to cover so has been split into 2 parts.  Let’s dive in. ~ Requirements The main requirements for the solution are: Detect motion Capture photo Create message with photo attached Send message to defined Telegram bot Detect who is in the photo, if the person is me, then do not invoke the Telegram bot with a message and image &nb

I was recently invited on a podcast with Grey Matter to share my experience, and thoughts from a development perspective. We discussed many topics.  Naturally AI was mentioned. Topics we discussed included: What drew me to a career in development? Which languages and tools  do I use now and why? How has development changed over the years since I started in the industry? What are the latest developer trends I’m seeing? How to stay up to date with the latest trends? Why and how has AI suddenly become such a big thing? How will AI affect developers and what they do? What challenges are developers facing now? What challenges could developers face in the future?   We also talk about Semantic Kernel and open-source.   Catch the episode here: Acast: -  https://shows.acast.com/the-grey-matter-podcast/episodes/developer-insights-and-ai Spotify: -  https://open.spotify.com/episode/6RgLQP2ytRIxXK6woFd0e2?si=LfmIwNAKRV6t_F_Y2GXtkg ~

I recently discovered I had been added to the Top 50 Global Thought Leaders and Influencers on Agentic AI for 2025 list on Thinkers360. The list is generated using a unique patented algorithm and takes a holistic measure of thought leadership and authentic influence looking that go beyond social media. It can help brands find the right experts for their niche. List members are curated by Thinkers360 for their thought leadership content. This can include:  articles  blogs  books  keynotes  media interviews panels podcasts social media speaking events videos webinars whitepapers   Thinkers360 leader boards are based on member’s real-world content and accomplishments. You can view the entire list of members that made it for 202 here.

I was looking to create a simple home security system.  The main requirements were being able to access a camera remotely available 24x7 low cost no subscriptions available via cell phone   After a look around, the Raspberry Pi Zero 2 and Zero Cam looked suitable. In this blog post, I document the software, hardware, pain points, and resolutions to build a camera stream that you can access remotely 24x7. ~ Hardware To build the surveillance solution, you need the following hardware: 1 x Raspberry Pi Zero 2 W - (£17) 1 x Zero Cam - Camera for Raspberry Pi Zero – (£15.00) 1 x Pibow Case for Raspberry Pi Zero 2W – (£6.00) 1 x Official Raspberry Pi Zero UK Power Supply – (£7.60) 1 x Raspberry Pi OS Pre-installed Micro SD Card 32GB - (£9.90)   Total cost for this solution: £55.50. ~ Software The following software is required: Raspberry Pi OS Rpicam VNC Viewer   The rpicam library may come installed with the OS. ~   Assembling and Testing the Camera As

Ive been working with Elastic Search recently. This post is more for my own refence when performing common tasks when creating code to integrate with Elastic Search using the Nest SDK. You might find it useful too. The following are covered: Creating an index Making a connection Adding a document to an Index Searching for documents Searching for documents using a parameterized field and value Updating a document Helpful REST API endpoints ~ Creating an Index How to create an index for storing documents: public async Task CreateIndexAsync(string indexName)         {             // check if index alread exists             var exists = await _elasticClient.Indices.GetAsync(indexName);                            Console.Writ

I like to encapsulate as many business rules and logic as possible in class libraries.   These are often used in web APIs, Azure Functions or wherever is required.   Sometimes you want to access values from an appsettings.json file that exists in a class library project but not pass values or JSON references through different layers of your solution to said class library.   This short post shows you how to do this.   Note: This is more for my own reference and when creating prototypes locally, but you may find it useful. A better approach for cloud solution configuration management is to use Azure Key Vault or Azure App Configuration. ~ Appsettings.json We have the following values in the JSON file: {  YourKey1: YourValue1,  YourKey2: YourValue2 }   Creating a Helper Class A static helper class ApplicationConfiguration is created. It imports the namespace Microsoft.Extensions.Configuration. The appsettings.json file is loaded from the directly the assembly is run

Build web apps and services that run on Windows, Linux, and macOS using C#, HTML, CSS, and JavaScript. Get started for free on Windows, Linux, or macOS.

If you are reading this blog, you probably listen to one or many podcasts. Some of my personal favourites include Seth Godin’s Akimbo, Lex Friedmans, Mike Tysons, and Andrew Hubermans. Most of these podcasts span multiple hours, and despite wanting to listen to or watch them, I just don’t have that amount of free time. ~ The Idea This got me thinking.  I could use AI to create a tool that would use speech-to-text to transcribe the audio.  I could then use AI to perform document summarisation on the transcription.   This would let me skim read the summarised points to get a feel for the main points, and outcome of the discussion points in the podcast.   I could use the summarisations to help me decide if the podcast was worth listening to. ~ CLI Tool In this blog post, you will see how I built a CLI tool using a combination of Azure AI services to save me hours each week.   Specifically, you learn about: Creating a CLI tool that accepts a podcast URL Batch transcription