Building AI with AI: Welcome to Sideslip Engineering

Introducing our new engineering blog where we'll share technical insights, architecture decisions, and the engineering culture at Sideslip.

By Michael Antczak
•
announcement

Welcome to the Sideslip Engineering Blog! We’re excited to launch this platform where our engineering team will share insights, experiences, and technical deep dives into how we build AI solutions using AI itself.

Why We’re Starting This Blog

At Sideslip, we’ve embraced a radical approach to AI development: we don’t just build AI solutions - we use cutting-edge AI tools to build them. While others debate whether AI will replace developers, we’re already using AI as our co-developer, architect, and code reviewer.

Think of it this way: a master craftsman uses the best tools available. When those tools happen to be AI models that can generate code, optimise algorithms, and catch bugs before they happen, why would we use anything else? This blog chronicles our journey at the intersection where AI builds AI.

What Makes Our Approach Different

We’re implementing Product as Code (PaC), a specification we’ve developed for defining software products in a structured, version-controlled, and AI-readable format. While others use AI as a coding assistant, we’re using it to bridge the entire gap from product definition to implementation.

Product as Code: Specification-Driven Development

Traditional development loses critical information between product documents, tickets, and code. PaC changes this by treating product definitions as structured data that both humans and AI can work with:

  • Version-controlled product definitions: Product specs live in Git, not lost in Confluence or PDFs
  • Structured YAML specifications: Clear schemas for features, user stories, and acceptance criteria
  • AI as the implementation engine: LLMs can read PaC specs and generate matching implementations
  • Traceable requirements: Every line of code traces back to a specific product decision

Here’s how we define products using the PaC specification:

# epic-01-user-authentication.yaml - AI can read this and build from it
apiVersion: productascode.org/v0.1.0
kind: Epic
metadata:
  id: 01HK8Q2M3NP7YBE4VTQZ8JF2RG
  sequence: 1
  name: user-authentication
  created_at: 2025-01-15T10:30:00Z
  updated_at: 2025-01-15T10:30:00Z
  started_at: 2025-01-16T09:00:00Z
  due_at: 2025-03-15T00:00:00Z
  completed_at: null
  labels:
    x-jira-epic: "AUTH-123"
spec:
  description:
    "Complete user authentication system with signup, login, and password
    recovery"
  status: in-progress
  priority: high
  owner: product-team
  success_metrics:
    - "90% signup completion rate"
    - "< 2 second login time"
    - "Zero critical security vulnerabilities"
  estimate:
    estimated:
      value: 21
      unit: "story_points"
      confidence: "medium"
    actual:
      value: 13
      unit: "story_points"
      confidence: "high"
  tickets:
    - "01HK8Q2N1QS9ZCF5WUXA9KG3SH"
    - "01HK8Q2P2RT0ADF6XVYB0LH4TI"
  epics:
    - "01HK8Q2Q3SU1BEG7YWZC1MI5UJ"
  labels:
    team: auth-team
    quarter: Q1-2025
    area: security
  related_to:
    - "01HK8Q2R4TV2CFH8ZXAD2NJ6VK"
    - "01HK8Q2S5UW3DGI9AYBE3OK7WL"

When we feed this specification to our AI toolchain, it doesn’t just generate code - it understands the product intent, creates architectural proposals, and maintains alignment between what we specified and what gets built.

From Specification to Solution

The PaC specification enables a new development workflow:

  • We use AI to transform business requirements into structured PaC specifications
  • AI validates completeness and suggests missing elements
  • Engineers and AI collaborate to implement the specification
  • Changes to requirements are tracked like code changes
  • AI ensures implementation matches specification continuously

This isn’t about replacing engineers - it’s about giving them and AI a common language. When both humans and AI can read the same specification, magic happens: development accelerates, misunderstandings disappear, and products actually match what stakeholders envisioned. The PaC specification is open source and evolving. We’re not just using it; we’re contributing to its development based on real-world AI development patterns we discover.

What You Can Expect

Our blog will cover a wide range of topics, including:

Technical Deep Dives

We’ll share detailed posts about our architecture decisions, performance optimizations, and the technical challenges we solve. Expect to see:

  • System design and architecture patterns
  • Performance optimization techniques
  • Scalability solutions
  • Security best practices

Engineering Culture

Building great products isn’t just about code – it’s about people. We’ll share insights into:

  • Our development workflows
  • How we approach code reviews
  • Team collaboration practices
  • Learning and growth opportunities

Open Source Contributions

We’re passionate about open source and will regularly share:

  • New projects we’re releasing
  • Contributions to existing projects
  • Guides for contributing to our repositories
  • Community highlights

Our Commitment to Quality

Every post on this blog will meet high standards:

  • Practical Examples: Real code and real solutions
  • Performance Metrics: Backed by data and benchmarks
  • Clear Explanations: Technical depth without unnecessary complexity
  • Interactive Demos: Where applicable, we’ll include live examples

Stay Connected

  • Follow us on X and LinkedIn for updates
  • Star our projects on GitHub
  • Subscribe to our RSS feed to never miss a post

We’re thrilled to embark on this journey of sharing and learning with you. Welcome to Sideslip Engineering – let’s build something amazing together!

Have questions or topics you’d like us to cover? Reach out to us at engineering@sideslip.io

← Back to all posts
Share on X