AI and .NET: a duo you may be underestimating
AI isn't just Python or JavaScript. If you're a .NET developer, you already have a complete ecosystem to integrate AI models into your projects, without switching languages.
If you think AI doesn't concern you because you write C#, because you lack time or skills, or because it's not your field... think again. Accessible tools and documentation already exist. Not getting started now means risking shooting yourself in the foot.
1. ML.NET: native machine learning in C#
Machine learning directly in C#, without going through Python. Model Builder and the ML.NET CLI let you build and deploy predictive models for web, desktop or cloud.
👉 The ML.NET getting-started tutorial
2. Azure AI Services: ready-to-use APIs
Add AI capabilities (vision, language, speech recognition) to your applications through Azure's pre-trained APIs, in a few lines of code.
👉 The Azure AI Services documentation
3. Semantic Kernel: AI agents connected to your data
Microsoft's open-source framework for integrating LLMs in .NET. Leverage language models, combine them with your internal services, APIs and databases to build AI agents able to orchestrate complex tasks.
It's exactly what I used to build a Pokémon mini-game in 20 minutes, with no database and almost no business logic.
👉 The Semantic Kernel documentation
4. Hugging Face: the Ali Baba's cave of AI
Thousands of pre-trained NLP and vision models, usable via API or locally, with tools to train and improve your own models, backed by a hyperactive community sharing models and datasets.
👉 The Hugging Face documentation
5. ONNX Runtime: run your AI models anywhere
Easily deploy models trained with TensorFlow or PyTorch in your .NET apps with ONNX Runtime. Fast on CPU and GPU, in the cloud or at the edge.
👉 The ONNX Runtime documentation
Where to start?
If you don't know where to begin, or want to discover what's possible with AI, go explore huggingface.co: interactive demos, ready-to-use models, APIs... a gold mine for experimenting and learning.
Takeaway: thanks to native tools and easily integrated external solutions, .NET lets you build AI-enriched applications without making your life complicated.
This article comes from one of my LinkedIn posts (in French): join the discussion ↗