Code smarter, not harder | .NET Day Switzerland 2025

Transcript

1. Code smarter, not harder How to AI Coding tools boost your productivity Daniel Sogl @sogldaniel Software architect
2. How to AI Coding tools boost your productivity Code smarter, not harder
3. Daniel Sogl • Software architect @ Thinktecture AG • MVP – Developer & Web Technologies • Focus: Angular and Generative AI • Socials: https://linktr.ee/daniel_sogl About me How to AI Coding tools boost your productivity Code smarter, not harder
4. The promises • Developers complete tasks 26% faster (MIT, 2023) • Coding tasks are on average 55% faster (GitHub, 2024) The (current) reality • AI-generated code with security vulnerabilities: 44% (Stanford, 2025) • Code duplicates increased by a factor of 4 after AI deployment (GitClear, 2024) Code smarter, not harder How to AI Coding tools boost your productivity Productivity claims over productivity claims https://mitsloan.mit.edu/ideas-made-to-matter/how-generative-ai-affects-highly-skilled-workers https://github.blog/news-insights/research/research-quantifying-github-copilots-impact-in-the-enterprise-with-accenture https://ee.stanford.edu/dan-boneh-and-team-find-relying-ai-more-likely-make-your-code-buggier https://www.gitclear.com/ai_assistant_code_quality_2025_research
5. How to AI Coding tools boost your productivity Code smarter, not harder
6. Code smarter, not harder How to AI Coding tools boost your productivity https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/
7. Code smarter, not harder How to AI Coding tools boost your productivity https://economictimes.indiatimes.com/magazines/panache/mit-study-shatters-ai-hype-95-of-generative-ai-projects-are-failing-sparking-tech-bubble-jitters/articleshow/123428252.cms
8. Code smarter, not harder How to AI Coding tools boost your productivity
9. Choosing the Right AI Tools for Your Development Workflow AI Tools for Developers: Prototyping vs. Coding How to AI Coding tools boost your productivity Code smarter, not harder
10. I want to create a web app that helps me generate social media posts. I should be able to authenticate myself to securely access my account. Once I’m logged in, I want to enter a topic into a free-text field. The app should then fetch current sources related to that topic using the Perplexity API. Based on these sources, it should use the ChatGPT API to generate relevant, engaging, and reach-optimized social media posts. All generated posts should be persistently saved in a database and linked to my authenticated user account. How to AI Coding tools boost your productivity Code smarter, not harder
11. In 5 minutes How to AI Coding tools boost your productivity Code smarter, not harder
12. Full-stack applications with a user-friendly approach • User friendly interface for nontechnical users • Uses React • Chat mode to plan before act • Strong Supabase integration for authentication and databases • 3rd Party integrations like Stripe, Resend or OpenAI • Optimized for team collaboration with GitHub- first workflows Tool Spotlight: Lovable.dev How to AI Coding tools boost your productivity Code smarter, not harder
13. How to AI Coding tools boost your productivity Code smarter, not harder
14. Code smarter, not harder How to AI Coding tools boost your productivity
15. Why coding tools fail to support you • LLMs were trained on older data • LLMs don’t understand your architecture without help • LLMs don’t know your company domains • LLMs don’t know your coding standards • Developers don’t write prompts describing every edge case or needed context to solve tasks • Different tasks require different context Code smarter, not harder How to AI Coding tools boost your productivity Limitations of LLMs
16. Code smarter, not harder How to AI Coding tools boost your productivity
17. Context. Context. Context. • We can define so-called custom instructions to provide extra context to all Copilot actions • A custom instruction is a markdown file with rules, instructions and guidelines defined by developers • Custom instructions can be added in a fine-grained way to specific file types • Custom instructions are part of a project and can be shared with all other developers • This pattern can also be used with Cursor, Windsurf, Claude Code or Gemini CLI • Demo: https://github.com/danielsogl/copilot-workflow-demo Info: There is currently no unified format supported by all tools - you need to create separate configuration files for each tool individually Code smarter, not harder How to AI Coding tools boost your productivity How to solve that problem?
18. Code smarter, not harder How to AI Coding tools boost your productivity Custom instructions example
19. Usage with Copilot • Copilot knows two types of instruction files • Global instructions: copilot-instructions.md • Fine-tuned instructions: custom.instructions.md • Instructions can be scoped so they are included for defined file types such as *.ts, *.html, *.css • This pattern also works with other tools like Cursor or Windsurf Tip: Keep your instructions short. Each condition will be part of your context window Code smarter, not harder How to AI Coding tools boost your productivity Custom instructions hierarchy
20. • Custom instructions can be generated based on your current project architecture and existing files • Let the AI tool analyze your project • Use an example instructions file as reference • Generate instructions for your architecture, dependencies, best practices and code style with the help of the AI tool, adding extra context – for example, your ESLint file • Use predefined instructions • https://codingrules.ai • https://angular.dev/ai/develop-with-ai • https://www.cursordirectory.com • https://github.com/github/awesome-copilot Code smarter, not harder How to AI Coding tools boost your productivity How to create custom instructions
21. • Determines how much code/context an LLM can handle simultaneously • Larger context → easier multi-file refactoring, whole-project reasoning • Example contexts: • Small (~4k tokens): Single file, short functions • Medium (16k–32k): Full modules, long functions • Large (100k+): Entire repositories, API docs, complex integrations Code smarter, not harder How to AI Coding tools boost your productivity Why Context Size Matters
22. Code smarter, not harder How to AI Coding tools boost your productivity
23. Fragmented AI Landscape: Multiple AI providers with different APIs, authentication and response types Complex Integration: Context management, tooling and response handling are challenging Security: Access control, 3rd party providers, private data can’t be accessed or shared Code smarter, not harder How to AI Coding tools boost your productivity Why MCP matters
24. The key for ”intelligent” workflows • Open-source protocol developed by Anthropic • Provides a consistent way for LLMs to interact with external resources • Client-Server architecture: AI applications (clients) request context from external services (servers) • Official servers are available for GitHub, Atlassian, Playwright, Stripe, Databases and more • It’s the key to useful AI-coding setups in complex environments Code smarter, not harder How to AI Coding tools boost your productivity Model Context Protocol (MCP)
25. Code smarter, not harder How to AI Coding tools boost your productivity MCP workflows Developer asks: “Explain failing tests in PR #42.” Copilot calls GitHub MCP → fetches PR diff + CI log LLM reviews context → returns root-cause & fix steps Dev clicks “Apply fix” → Copilot edits code & opens new PR
26. • Microsoft verified MCP Servers for Agent Mode: https://code.visualstudio.com/mcp • MCP GitHub Repo: https://github.com/modelcontextprotocol/servers • Awesome MCP Server: https://github.com/punkpeye/awesome-mcp-servers • MCP Collection: https://glama.ai/mcp • Official provider websites like Atlassian, GitHub or PayPal Code smarter, not harder How to AI Coding tools boost your productivity How to find MCP servers?
27. Code smarter, not harder How to AI Coding tools boost your productivity The best developers won’t be replaced by AI, they’ll be the ones who learn to wield it
28. Thanks for your attention Daniel Sogl @sogldaniel daniel.sogl@thinktecture.com