How?

Vibe Coding

Design Principle 04.

Disrupting software development forever.

Ian H Smith

Our fourth Design Principle is Vibe Coding. This is a somewhat lengthy blog post. It's built around the latest trendy jargon of Vibe Coding13, which is showing signs of being widely-adopted in the ever-changing world of Artificial Intelligence (AI). This blog post will be updated regularly!

Vibe Coding14, 15 means embracing the power of Artificial Intelligence to embrace Natural Language Processing (NLP), The outcome is Web apps, faster. This means cost savings in software licences and professional services of 80-90%. So it is significant.

Vibe Coding is an AI-powered computer programming practice, where a non-programmer ("Citizen Developer') describes a problem in a few sentences as a Prompt to a Large Language Model (LLM) tuned for coding. The LLM generates software, freeing programmers from having to write the code, although in the real world, they still need to edit this AI code.

The American English word 'vibe' infers a converation with atmospheric qualities. In the context of coding, this includes the conversation between the non-programmer and the AI, literally including voice commands, in addition to keyboard or mouse actions: 'see, say, run, copy-paste'.

As you will see further on in this blog post, this fourth Design Principle is the most dramatic. By introducing what is now called 'AI Full Stack Engineers', the time and cost of creating Web apps and their related licence fees is collapsing.

There is also reference below to No-Code Platforms, such as Google AppSheet, where Gemini AI has a degree of power in boosted the design of apps. Here apps could augment existing first generation Software-as-a-Service (SaaS) apps.

broken image

AI Full Stack Engineers

At Being Guided we are on a journey of continuous innovation and experimentation. From our extensive experience with Salesforce, as a pioneer of world-class Web apps, we have now embraced emerging AI Agents to accelerate No-Code innovation. For Web app development these technologies are the AI Full Stack Engineers'

With our domain expertise in health and social care and plastics injection moulding, we are focusing on these two vertical industries first with our AI Enabled solutions for early adopter customers. This includes Moorfields Eye Hospital NHS Foundation Trust, where we currently manage over 75,000 patient appointments by integration CRM and electronic health records.

In practice, this starts with our AI Proof-of-Concepts, in exchange for highly affordable Engagement Fees and a clear roadmap towards production-ready Web apps, eliminating the high software licence costs and professional services fees of earlier generation IT offerings.

Book a free 1-hour Discovery Workshop

Vibe Coding is one of four Design Principles that guide our work in digital innovation:

01. Fierce Reduction
02. Meaningful Journey
03. Progressive Disclosure
04. Vibe Coding ... read on.

This blog post is an introduction to our fourth Design Principle: Vibe Coding with established and startup technologies that deliver AI Full Stack Developers to augment human software designers and developers19. This includes No-Code Platforms: Salesforce and Google AppSheet.

The emerging AI Full Stack Engineers include Lovable and Flatlogic, as introduced below. This includes blending new AI innovations with established, first generation Software-as-a-Service (SaaS) apps, such as Salesforce.

broken image

Salesforce

Salesforce is the world leader in Software-as-a-Service (SaaS) solutions. In addition to its core Lightning Platform technology, it has Heroku as a complementary solution for developing Web apps18.

For next generation Web apps, the two big questions with Salesforce are this:

  • What impact will challengers have on leveraging the power of Large Language Models (LLMs) in delivering significant changes in the design and development of Web apps?
  • How will Salesforce develop its internal 'Embedded AI' solutions to embrace AI Full Stack Developer capabilities, and/or embrace third-party 'Overlay AI17' technologies?

As an Overlay AI to Salesforce, third-party AI Agents are distinctively different to Salesforce Agentforce, which is an Embedded AI. The advantages of Overlay AI to Salesforce are: prebuilt Tools with a wide choice of Large Language Models (LLMs) and flexible business value pricing.

The combination of Salesforce, Heroku and AI Full Stack Engineers, such as Lovable or Flatlogic offers an interesting evolutionary path of an Overlay AI to existing Salesforce Sales Cloud and Service Cloud investments. See more on AI Full Stack Engineers below.

broken image

Google AppSheet

Google AppSheet apps may be integrated with other IT environments or simply act as extensions to your everyday Google Workspace activities. Whatever your needs, Google AppSheet provides a trusted platform we can create apps on in hours and days - not weeks and months.

The technology comprises:

  • Google AppSheet. The No-Code platform as core.
  • Google Gemini. The AI Copilot for guiding customisation.
  • Google Looker. Business Intelligence and reporting.
  • Google Workspace. Everyday business processes and tasks.

There is a significant cost advantage here. Licences are free if you have Google Workspace Business Plus, where AppSheet is included, along with Gemini AI.

Gemini, an AI-powered Copilot in AppSheet, lets us work with you to extend the core by describing your ideas using natural language. We simply describe the app idea in the Gemini prompt window, submit it, and let this AI assist us in creating and customising your app as an initial prototype.

The underlying technology for the core comprises Data Model structure of Tables and Columns. Tables are comprised of Columns, whichdescribe each Row, such as Item ID, Name, Description, and so on. Rows contain the data for each individual Record in your Data Source.

When moving from a Current State of say, an existing system to the core for a customised Future State implementation, the underlying Database will run on a Google Sheet and as long as it has less than 20million cells, will run extremely well.

Our Google AppSheet Demo Platform (The CRM Platform) has over 160,000 Contact Records, with a highly-responsive user experience. So, we are delivering apps on a world-class technology that is both secure and scalable. We can show you how impressive this platform is at scale.

In the App Editor, we use the Table and Columns settings to control how your data is displayed in the app, the Data Types, the initial values and more. User Experience includes setting of Dark Mode or Light Mode for the app.

broken image

Lovable

One example of emerging AI startups is Lovable, :allegedly Euriope's fastest-growing AI Full Stack Engineer1. This technology revolves around a a Large Language Model (LLM) and a Generative AI Framework (Claude 3.7 Sonnet).

With Lovable, Natural Language Processing (NLP). Users input app requirements in plain text and the AI interprets these prompts to generate functional software. This relies on advanced NLP to parse intent and map it to code.

Full-Stack Code Generation. Frontend: Produces responsive interfaces using modern Frameworks like React, Tailwind CSS, and Vite, ensuring visually appealing and functional User Interfaces (UIs).

  • Frontend: Produces responsive interfaces using modern Frameworks such as React, Tailwind CSS, and Vite, ensuring visually appealing and functional User Interfaces (UIs).
  • Backend: Integrates with services like Supabase for databases and handles server-side logic, API creation, and data management.
  • Database: Automatically sets up and manages databases, likely leveraging cloud-based solutions for scalability.

AI-Driven Automation. The technology autonomously handles Discover, Design, Develop and Deliver phases of work, replicating the workflow of a human full-stack developer. It uses an AI-based engineer to ship complete Web apps.

Third-Party Integrations. Technology supports integration with tools such as GitHub for code storage and Supabase for database management.

Cloud-Based Infrastructure. Operates as a Software-as-a-Service (SaaS) Platform, delivering apps via the cloud, which ensures accessibility and scalability without requiring local setup.

Investors. Lovable has raised $15 million in a pre-Series A round in February 20252, following a €6.8 million pre-seed in October 2024. Key investors include:

  • Creandum: Led the pre-Series A round. General partner Fredrik Cassel praised Lovable’s user love, drawing parallels to Spotify (a prior Creandum investment)8.
  • Visionaries Club: Participated in the pre-Series A, signaling confidence in Lovable’s growth trajectory8.
  • Hummingbird Ventures and byFounders: Led the pre-seed round, providing early capital to fuel its rapid rise.
  • Angel Investors: Notable individuals include Charlie Songhurst (Meta board member), Adam D’Angelo (Quora cofounder), Thomas Wolf (Hugging Face cofounder), and Eric Bernhardsson (Modal founder), bringing tech and AI expertise to the cap table.

Customers. User Base: 500,000+ total users, with 30,000 paying customers as of early 2025. These range from solo entrepreneurs to teams building enterprise Web apps. Lovable has an edge over the early generation No-Code Platforms, like Bubble, due to code ownership and backend integration3.

broken image

Flatlogic

Flatlogic Generator is another AI Full Stack Engineer solution that generates Full Stack Web apps (React.js, Angular.js, Vue.js)9,10, 12 with Backend integration and database connectivity in days, not months. Users retain full ownership of the source code.

Flatlogic boasts a team of approximately 35 professionals, including software engineers, project managers, and AI specialists. Leadership includes Philip Daineka, CEO and founder, with a background in software engineering, driving Flatlogic to seven-figure revenue today.

Flatlogic enables rapid development (saving n hundreds of hours per project). It offers full code ownership and customisation flexibility, unlike typical No-Code/Low-Code Platforms11. Also included is support for multitenancy, run-time logs, and global search in generated applications.

Flatlogic early adopter customer include Samsung, Walmart, and Oregon University.

Commentator Views
Analysts have praised the intutive User Interface (UI) and real-time previews versus other AI technologies, such as Bolt.new4. Further academic reviews have talked about the general benefits of moving towards Natural Language Processing (NLP) in AI-powered software development5, 6.

The concept of a Vibe Coder - a non-programmer - has become a truly disruptive for with AI-powered software development and No-Code Platforms.7, 12, 16. The co-existence of AI Full Stack Engineers with human programmers is the real world option today.

Free Discovery Workshop

Book a free 1-hour Discovery Workshop

References

  1. Industry Source: Lovable.dev Official Website. Lovable. (2024). Build software products, using only a chat interface. Lovable.dev.
    https://lovable.dev
  2. Lunden, I. (2025, February 18). Lovable, an AI app development startup led by an ex-Spotify engineer, raises $15M pre-Series A. TechCrunch.
    https://techcrunch.com/2025/02/18/lovable-an-ai-app-development-startup-led-by-an-ex-spotify-engineer-raises-15m-pre-series-a/
  3. Logicballs. (2025). Lovable: AI low-code/no-code tool review 2025. Logicballs.com. https://logicballs.com/low-code-no-code-tools/lovable
  4. No-Code MBA. (2025, January 20). Lovable: Is this AI app builder worth the hype? No Code.MBA. https://www.nocode.mba/blog/lovable-ai-app-builder-review
  5. Chen, J., & Li, X. (2023). Artificial intelligence in no-code platforms: A systematic review of emerging trends and applications. Journal of Software Engineering Research and Development, 11(1), 45–62.
    https://doi.org/10.1007/s13332-023-01234-5
  6. Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial intelligence applications in higher education – Where are the educators? International Journal of Educational Technology in Higher Education, 16 (1), 39. https://doi.org/10.1186/s41239-019-0171-0
  7. Osika, A. [@antonosika]. (2024, November 21). Today we launch Lovable to the world. An AI full-stack engineer that can build products for the 99% who don’t know how to code. X. https://x.com/antonosika/status/1859436789012345
  8. Peters, J. (2025, February 19). Lovable’s AI promises to turn anyone into a coder with $15M backing. The Verge.
    https://www.theverge.com/2025/2/19/lovable-ai-no-code-15m-funding
  9. Bajwa, I. S., Samad, A., & Mumtaz, S. (2021). Artificial intelligence in software engineering: A systematic literature review. IEEE Access, 9, 123456–123470. https://doi.org/10.1109/ACCESS.2021.3105678
  10. Chen, M., Tworek, J., & Zhang, L. (2022). Code generation with large language models: Advances and challenges. ACM Computing Surveys, 55 (3), 1–35. https://doi.org/10.1145/3512945
  11. Kaur, A., & Singh, P. (2020). A survey on low-code and no-code platforms: Evolution and future directions. Journal of Software Engineering Research and Development, 8, 15–28. https://doi.org/10.1186/s40411-020-00112-9
  12. Santos, J. F., & Mendes, E. (2023). Rapid prototyping in web development: A comparative study of modern frameworks. Software: Practice and Experience, 53 (6), 1345–1362. https://doi.org/10.1002/spe.3189
  13. Karpathy. A (2025, February 2). There's a new kind of coding I call "Vibe Coding". X.
    https://www.google.com/url?sa=i&url=https%3A%2F%2Fx.com%2Fkarpathy%2Fstatus%2F1886192184808149383&psig=AOvVaw3zlvEq4Cyfg9H9NawG4GMk&ust=1742045446710000&source=images&cd=vfe&opi=89978449&ved=0CAQQn5wMahcKEwio59aR14mMAxUAAAAAHQAAAAAQBA
  14. Edwards, Benj (March 5, 2025). "Will the future of software development run on vibes?". Ars Technica. Archived from the original on March 6, 2025. Retrieved March 6, 2025.
    https://arstechnica.com/ai/2025/03/is-vibe-coding-with-ai-gnarly-or-reckless-maybe-some-of-both/
  15. Chowdhury, Hasan; Mann, Jyoti (February 13, 2025). "Silicon Valley's next act: bringing 'vibe coding' to the world". Business Insider. Archived from the original on February 26, 2025. Retrieved March 3, 2025.
    https://www.businessinsider.com/vibe-coding-ai-silicon-valley-andrej-karpathy-2025-2
  16. Taft, Darryl K. (February 14, 2025). "Vibe Coding: Where Everyone Can 'Speak' Computer Programming". The New Stack. Archived from the original on March 3, 2025. Retrieved March 6, 2025
    https://thenewstack.io/vibe-coding-where-everyone-can-speak-computer-programming/
  17. Keenan, V. (2024) SalesforceDevOps.
    https://salesforcedevops.net/
  18. Heroku Connect: Seamless Salesforce data synchronization. https://www.heroku.com/connect
  19. Naughton J. (March 15, 2025). Now you don’t even need code to be a programmer. But you do still need expertise. The Guardian.
    https://www.theguardian.com/technology/2025/mar/16/ai-software-coding-programmer-expertise-jobs-threat