Digital Product & AI-Assisted Development2023 - Present

Digital Product & AI-Assisted Software Development

From process logic to code-based development, using AI to accelerate implementation

RoleSolutions Consultant & Developer
Organization / ContextConfidential & Personal Projects
Anonymization & Confidentiality NoticeThis case brings together practices and lessons applied across different digital projects, including work for third parties and personal projects. Client names, brands, domains, and proprietary business logic have been omitted or aggregated to preserve confidentiality.

Overview

Evolution from building no-code solutions to developing digital products with code and intensive AI assistance. I define and lead the business logic, architecture, flows, data modeling, security criteria, and validation, while AI accelerates the writing, review, and evolution of the code.

The Technical Challenge

As projects became more complex, no-code tools were no longer sufficient for some requirements. The challenge became building code-based solutions without losing control over business rules, security, data, integrations, and implementation quality.

  • Business rules and flows that required greater control than no-code platforms could provide.
  • Multiple user roles, permissions, and data-security policies.
  • External integrations, automations, and failure scenarios requiring explicit handling.
  • Need to evolve quickly without compromising testing, data integrity, or maintainability.

The Solution Blueprint

Adopted AI-assisted code development, combining functional and architectural ownership with intensive use of AI to accelerate implementation, testing, refactoring, documentation, and problem investigation.

  • Defined architecture, flows, business rules, data models, and security criteria before implementation.
  • Modeled relational data in Supabase / PostgreSQL with Row-Level Security (RLS), referential integrity, and concurrency control when required.
  • Built integrations and automations using APIs, webhooks, and n8n, including explicit exception and failure handling.
  • Used AI intensively for code writing and refactoring, test generation, documentation, and bug investigation.

Results & Evidence

No-code → CodeTechnical evolutionExpanded ability to build digital products
AI-AssistedDevelopmentAI as an accelerator for implementation and code evolution
Tests + RLSReliabilityValidation, data security, and access control

Impact & Operational Outcomes

  • Expanded ability to build products beyond the limitations of no-code platforms.
  • Greater control over architecture, data, security, integrations, and business rules.
  • AI used as an implementation accelerator while decision-making, review, and validation remain under human responsibility.

Development Approach

Development guided by product logic and business rules, using AI as an implementation and technical-support tool rather than a substitute for decision-making and validation.

Technologies & Methods
AI-Assisted DevelopmentSupabase / PostgreSQLRow-Level Security (RLS)n8nREST APIs / WebhooksRelational ModelingAutomated Testing
Core Engineering Principles
  • AI accelerates code writing; architecture, rules, criteria, and validation remain under my responsibility.
  • Understand the problem and define the logic before implementation.
  • Security, data integrity, exception handling, and testing are part of the solution, not afterthoughts.