AI-Assisted SE Guide

Appearance

About the Vibe Coding Guide ​

TLDR ​

The Vibe Coding Guide is a comprehensive resource for AI-assisted software development, created through a collaboration between The Low Code Foundation (a French non-profit) and Visual Hive (an AI startup based in London). Written by Richard Osborne, CTO of Visual Hive, this guide documents proven methodologies for building production-ready applications with AI.

The Collaboration ​

The Low Code Foundation ​

The Low Code Foundation is a French non-profit organisation dedicated to:

  • Promoting low-code development — Making software development accessible to more people
  • Training and education — Providing resources and training for modern development approaches
  • Community building — Connecting developers, businesses, and educators around efficient development practices

The foundation believes that the future of software development lies in reducing unnecessary complexity while maintaining quality and capability.

Visual Hive ​

Visual Hive is an AI startup based in London focused on:

  • AI-assisted conference tools — Building software that helps conference organisers increase attendee and exhibitor satisfaction, as well as overall revenue
  • Data analytics — Using data collected from conference deployments, tools and partners to create critical data insights
  • Managed services — Creating custom solutions for a wide range of problems in the conference industry and beyond

Visual Hive's mission is to make AI a genuine multiplier for conference organisers, not just a novelty.

The Author ​

Richard Osborne is the CTO of Visual Hive and the primary author of this guide.

Drawing from hands-on experience building production applications using AI-assisted development, Richard has documented the patterns, practices, and pitfalls discovered through real projects—including the RISE project that serves as the primary example throughout this guide.

The Methodology ​

This guide exists because most approaches to AI-assisted coding fail. People expect to describe an app and have AI build it. The reality is more nuanced.

The methodology documented here was developed through practical experience:

  1. Validate first — Use AI for exploratory conversations before committing to build
  2. Document systematically — Create lean documentation that gives AI the context it needs
  3. Work in focused tasks — New conversation per task, with clear scope
  4. Check quality continuously — Confidence scoring and phase audits catch problems early
  5. Build incrementally — Each phase must work before starting the next

This isn't theory. It's what actually works.

Open Source ​

This guide is open source and available on GitHub.

Contributions, corrections, and improvements are welcome. If you've found patterns that work or pitfalls to avoid, the community benefits from your experience.

Contact ​

Work with Visual Hive

Need help putting this methodology into practice? Visual Hive offers free consultations for teams and founders looking to adopt AI-assisted development — whether that's a strategy audit to understand where AI fits in your workflow, hands-on support building your first AI-assisted project, or full-service development and deployment.

Book a Free Consultation →Visit visualhive.co

Acknowledgments ​

This guide was built using the methodology it documents—AI-assisted development with human oversight, quality gates, and iterative refinement. The irony is intentional. The meta-ness is a feature, not a bug.

Special thanks to the developers, founders, and teams who tested these approaches and provided feedback that shaped the methodology.

"Trust the process. It works."