Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Code smarter, not harder | BASTA Main 2025

Avatar for Daniel Sogl Daniel Sogl
September 23, 2025

Code smarter, not harder | BASTA Main 2025

Software development is no longer just about writing code – it’s about efficiency, smart solutions, and focusing on what truly matters. AI-powered coding tools like GitHub Copilot, Cursor, bolt.new, and v0 are transforming the way we build software. This talk provides a comprehensive overview of the tools currently available, their use cases, and their limitations. It explores how these tools automate repetitive tasks, accelerate development processes, and create space for more creative and strategic work. Challenges and limitations are also addressed to provide a realistic perspective on their potential. The goal of this talk is to demonstrate how AI coding tools can optimize workflows and make day-to-day work more productive – without compromising on quality.

Avatar for Daniel Sogl

Daniel Sogl

September 23, 2025
Tweet

More Decks by Daniel Sogl

Other Decks in Programming

Transcript

  1. Code smarter, not harder How to AI Coding tools boost

    your productivity Daniel Sogl @sogldaniel Software architect
  2. Daniel Sogl • Software architect @ Thinktecture AG • MVP

    – Developer & Web Technologies • Focus: Angular and Generative AI • Socials: https://linktr.ee/daniel_sogl Code smarter, not harder About me How to AI Coding tools boost your productivity
  3. Democratization of Software Development • Software development is now accessible

    to everyone – powered by natural language. • GitHub Copilot: 50,000+ organizations • Path to "a billion programmers” - GitHub CEO Code smarter, not harder The AI Coding Revolution https://www.elitebrains.com/blog/aI-generated-code-statistics-2025 https://www.netcorpsoftwaredevelopment.com/blog/ai-generated-code-statistics https://blog.github.com How to AI Coding tools boost your productivity
  4. • GitHub Copilot: 55% faster task completion • Amazon CodeWhisperer:

    57% speed improvements • Cursor users report: 5-10x productivity multipliers • 96% of developers use Copilot the same day they install it Code smarter, not harder The Promise Machine https://github.blog https://aws.amazon.com/q/developer/ https://cursor.com/students How to AI Coding tools boost your productivity
  5. Controlled Study: 16 Experienced Developers, 246 Real Tasks • Actual

    performance: 19% SLOWER with AI tools • Developer perception: Believed they were 20% faster • Perception gap: 40 percentage points difference • Tool used: Cursor with Claude Sonnet 3.5 Code smarter, not harder Reality Check - The METR Study https://metr.org/blog/2025-07-10-early-2025-ai-experienced-os-dev-study/ How to AI Coding tools boost your productivity
  6. GitClear Analysis: 153-211 Million Lines of Code • Code churn:

    Projected to double in 2025 • Copy-pasted code: 8.3% → 12.3% (+48%) • Code refactoring: 25% → <10% (-60%) Code smarter, not harder Code Quality Catastrophe https://www.gitclear.com/ai_assistant_code_quality_2025_research How to AI Coding tools boost your productivity
  7. Veracode 2025 Study: AI Code Security Failures • 45% of

    AI-generated code fails security tests • Java security failure rate: 72% • Secret leakage: 40% higher in Copilot repos Code smarter, not harder Security Nightmare https://www.veracode.com/blog/genai-code-security-report/ https://blog.gitguardian.com/yes-github-copilot-can-leak-secrets/ https://snyk.io/reports/ai-code-security/ How to AI Coding tools boost your productivity
  8. The methodology gap that's killing productivity • The tools work

    for simple, isolated tasks • They fail in complex, real-world codebases because they lack context • Solution: Stop "vibe coding" and start strategic AI integration • Next: How to give AI the context it needs to succeed Code smarter, not harder The Real Problem Isn't AI - It's How We Use It How to AI Coding tools boost your productivity
  9. Choosing the Right AI Tools for Your Development Workflow Code

    smarter, not harder AI Tools for Developers: Prototyping vs. Coding How to AI Coding tools boost your productivity
  10. How AI can help you try out ideas in minutes

    • Try out ideas with simple prompts • Incorporate existing design systems using Figma or screenshots • Create full-stack applications with backend functionality like authentication, storage, and serverless functions Code smarter, not harder Agentic Prototyping: AI-Driven PoC’s How to AI Coding tools boost your productivity
  11. 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. Code smarter, not harder How to AI Coding tools boost your productivity
  12. In 10 minutes Code smarter, not harder How to AI

    Coding tools boost your productivity
  13. 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 Code smarter, not harder Tool Spotlight: Lovable.dev How to AI Coding tools boost your productivity
  14. Self hosted full-stack application prototyping • Based on bolt.new •

    Runs locally on your device using Docker • In browser Visual Studio Code environment • Can connect to any LLM provider or local running LLMs with Ollama • Supports any web framework or Node based backends Code smarter, not harder Tool Spotlight: bolt.diy How to AI Coding tools boost your productivity
  15. The Benefits and Limitations of AI-Assisted Development • Fast PoC

    development – Reduces time from idea to prototype • No Local Setup Needed – Tolls running in the browser • Immediate Visual Feedback – preview UI changes in real-time • Supabase Integration – Strong backend setup in Lovable • Collaboration Features – Lovable integrates well with GitHub • AI costs and token limits – usage is often billed per token or message • AI-generated code requires manual refinement – quality varies depending on prompts and models. • Vendor lock-in – code can be shared between tools • Limited framework support (mostly React) Code smarter, not harder AI-Driven Prototyping: Pros & Cons Pros Cons How to AI Coding tools boost your productivity
  16. How AI Automates, Refactors, and Optimizes Code • Add tools

    by plugins or use an AI-Coding IDE • Pretend architecture & tech stack details by providing rules • Use different LLM models and begin prompting • Use images & documentations to add current knowledge • Add external context using MCP-Servers • Use the agent mode or the chat to plan or act • Develop features or refine existing code as needed Code smarter, not harder From Prototype to Production: AI Coding Agents in Action How to AI Coding tools boost your productivity
  17. The all-inclusive AI Coding toll • Can be used accross

    multiple IDEs (VS-Code, IntelliJ, Xcode etc.) • Free to use / 10$ per month for the pro version • Provides ask, edit and agent mode • Inline autocompletion, code reviews and test generation • Povides different LLMs (GPT, Gemini, Sonnet, etc.) • MCP (Model Context Protocol) support Code smarter, not harder Tool Spotlight: GitHub Copilot – Edit/Agent Mode How to AI Coding tools boost your productivity
  18. 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 Limitations of LLMs How to AI Coding tools boost your productivity
  19. Simplify your prompts • Project-specific coding standards • Architectural patterns

    and constraints • Domain knowledge embedding • Step-by-step reasoning requirements How to generate?: • Ask the AI to analyze your codebase • Use predefined instructions • Iterateive update your instructions Code smarter, not harder Fighting Back - Custom Instructions How to AI Coding tools boost your productivity
  20. AGENTS.md • The problem: each tool uses different file names

    for custom instructions • AGENTS.md aims for becoming supported by all AI tools • Supported by: Cursor, GitHub Copilot, Aider and more Code smarter, not harder Standardization of Custom Instructions How to AI Coding tools boost your productivity
  21. Override the default System Prompt • Create agents for specific

    tasks • Define supported tools, MCP-Servers and LLM- Models • Run reusable prompts for reusable workflows How to AI Coding tools boost your productivity Code smarter, not harder Custom Chat Modes
  22. 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 Model Context Protocol (MCP) How to AI Coding tools boost your productivity
  23. 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 Code smarter, not harder MCP workflows How to AI Coding tools boost your productivity
  24. • Not all LLMs are equal: Performance varies widely •

    Understand reasoning vs. non-reasoning models • Context size matters greatly in coding scenarios • Benchmarks help guide practical model choice • Cost per 1 million input/output tokens Code smarter, not harder Choosing the right LLM for your tasks How to AI Coding tools boost your productivity
  25. Choose wisely Code smarter, not harder Popular LLMs for software

    development Model Provider Reasoning Context Window SWE-bench $ / 1M tokens Claude 4 Opus Anthropic 200k tokens 72.5 % $15 / $75 Claude 4 Sonnet Anthropic 200k tokens 72.7 % $3 / $15 GPT-5 (high) OpenAI 128k tokens 74.9 % $1.25 / $10 GPT-o3-mini (high) OpenAI 200k tokens 49.3 % $1.10 / $4.40 Gemini 2.5 Pro Google 1M tokens 63 % $1.25 / $10 DeepSeek V3 DeepSeek 163k tokens 38.3 % $0.25 / $0.85 Grok 4 xAI 256k tokens 72 % $3 / $15 Kimi K2 Moonshot 131k tokens 65.8 % $0.13 / $0.13 How to AI Coding tools boost your productivity
  26. CO-STAR Framework for Prompting • Context: Project background, constraints •

    Objective: Specific goals, success criteria • Style: Coding patterns, conventions • Tone: Collaborative, step-by-step • Audience: Experience level, domain • Response: Format, structure requirements Code smarter, not harder Prompt Engineering That Works How to AI Coding tools boost your productivity
  27. Proven Tools Stay Essential • Static Analysis Foundation: ESLint, SonarQube,

    SAST tools remain critical • AI-Enhanced Review: CodeRabbit, Copilot suggestions as additional layer • Automated Security Scanning: Snyk, Veracode for vulnerability detection • Comprehensive Testing: Unit, integration, security tests (AI can generate, not replace) • Human Oversight: Architectural decisions, business logic validation Code smarter, not harder Defense in Depth: Quality Gates for AI-Enhanced Development How to AI Coding tools boost your productivity
  28. How to Do AI Coding Right Setup: • Define custom

    instructions/rules files • Configure MCP servers for your stack • Choose task-appropriate LLMs Process: • Use CO-STAR prompt framework • Require AI to explain before coding • Implement automated testing generation • Maintain human review of critical paths Quality: • Run security scans on all AI code • Code quality: Cyclomatic complexity, test coverage, refactoring frequency • Track technical debt metrics Code smarter, not harder Best Practices Checklist How to AI Coding tools boost your productivity
  29. Your Next Steps Experiment Responsibly • Start with low-risk projects

    • Implement quality safeguards first • Measure Everything Track actual productivity (not perception) • Monitor code quality metrics • Assess security vulnerabilities Invest in Skills • Prompt engineering techniques • AI tool orchestration • Quality assurance processes Code smarter, not harder Call to Action How to AI Coding tools boost your productivity
  30. Code smarter, not harder The best developers won’t be replaced

    by AI, they’ll be the ones who learn to wield it How to AI Coding tools boost your productivity