A Thinking Model for Data Architecture
Vellamon Kalliot is a thinking model drawn directly from over 25+ years of real-world experience — both the successes and the valuable lessons from trials and errors. It is not something newly invented, but rather an extraction of what consistently works in practice.
This gives you a reliable, always-working framework that helps you navigate the complex world of software engineering with clarity and confidence. With thousands of methodologies, standards, and principles out there, it’s completely understandable to feel overwhelmed at times. Vellamon Kalliot cuts through the noise and offers a simple, practical mental model to understand and master the chaos.
Vellamon Kalliot (The Cliffs of Vellamo) is a thinking model for data architecture. It provides a simple, systematic way to approach the design and long-term management of data systems.
Rather than focusing on specific tools or modeling techniques, Vellamon Kalliot emphasizes fundamental structural principles. It helps architects, developers, and organizations build data systems that remain understandable, trustworthy, and maintainable — even as technologies and business requirements change rapidly.
The model is particularly relevant in environments with increasing automation and AI, where multiple agents interact with data at high speed.
Who is Vellamo?
Vellamo (also spelled Wellamo) is a figure from Finnish mythology. She is known as the goddess of water, lakes, and seas — often referred to as Veen emäntä, meaning “Mistress of Water.”
In traditional stories, Vellamo governs the movement and power of water. She is associated with waves, depth, and the unpredictable yet vital nature of water. She is frequently paired with Ahti, the god of the sea.
Why “Vellamon Kalliot”?
The name Vellamon Kalliot (“The Cliffs of Vellamo”) draws from this mythological and natural imagery.
In Finland, granite cliffs are among the most defining and enduring features of the landscape. They shape how water behaves — forming rivers, containing lakes, and creating coastlines. Without these solid structures, water would remain formless and uncontrolled.
This relationship forms the central metaphor of the model:
- Data is like water — fluid, powerful, and difficult to manage without proper boundaries.
- The Cliffs of Vellamo represent the solid, enduring structures that give data shape, direction, and protection.
Core Metaphor: Data as Water
Vellamon Kalliot uses this natural metaphor to make complex architectural ideas more intuitive:
- Water on its own is difficult to contain and direct.
- When shaped by cliffs and basins, water becomes useful — forming rivers, lakes, and seas.
- On high ground, water can freeze into ice, becoming stable and long-lasting.
This mirrors how data behaves in modern systems: without clear structure, data remains chaotic and hard to govern. With proper boundaries and classification, data becomes manageable, traceable, and reliable over time.
The Two Mountains
A central visual representation of the model contrasts two opposing structures:
- The Granite Mountain (solid, stable, right-side-up) represents the core data architecture — the part of the system that should remain trustworthy and long-lasting. Like information carved into granite, it is meant to endure and survive any catastrophe, providing a stable foundation even when everything else changes.
- The Cloud Mountain (fluid, inverted, spinning) represents the fast-changing and fluid layer needed to respond to constantly evolving client demands, shifting requirements, dynamic environments, and the demands of the AI era.
The model argues that many modern systems have been built upside down — treating the fragile, rapidly changing upper layer as the foundation while forcing the data layer to constantly adapt to it. Vellamon Kalliot instead proposes a clear separation: allow the Cloud Mountain to evolve freely and at the required speed, while protecting the Granite Mountain as the stable foundation that must remain intact.
Foundational Principles
Vellamon Kalliot is built on three core principles. These principles are interconnected, with Accountability serving as the foundation.
Accountability
The principle of accountability originates from accounting practices, where the primary goal is to ensure that liabilities and obligations are properly recorded, tracked, and can be demonstrated when required.
In data architecture, accountability serves a similar purpose: it ensures that data remains correct and that responsibility for it is clear. This includes both clear ownership and proper structural organisation of data. By treating data with the same rigor traditionally applied to financial liabilities, the model helps maintain integrity, traceability, and trust in complex systems. It is the primary principle of Vellamon Kalliot.
Demonstrability
Demonstrability is the ability to trace, explain, and prove what happened to data over time. It depends directly on strong accountability.
Operability
Operability refers to the ease of running, maintaining, and evolving systems long-term. It becomes achievable when accountability and demonstrability are properly addressed.
Key Structural Concepts
The model encourages architects to classify data and system components according to their fundamental nature. The most important distinctions are:
Functionality vs Persistence
Recognize the natural tension between the changing application layer and the need for stable data storage.
Mutable vs Immutable
Clearly distinguish data that is allowed to change from data that should remain stable once committed.
Immutable data is like information carved into granite — it is meant to outlive any catastrophe and remain trustworthy for a very long time.
Data vs Logs
- Data represents valuable, long-term information that should generally be preserved and protected.
- Logs represent records of events or actions. While important for traceability and auditability, logs often lose their relevance over time and can eventually be purged. Logs are like ice — frozen snapshots of flowing information. Hold them while they're useful, then let them melt back into the stream and disappear.
Log Usage Example for GDPR
In an order system, information such as the customer name and customer ID is considered Log data under the Vellamon Kalliot approach.
Rather than storing these as fields inside the main order record, they should be kept as separate log records (for example in a dedicated log table or as structured JSON entries). This separation makes it much easier to erase or anonymize personal data when required by regulations such as GDPR, without affecting the core business event.
The order event itself (that an order was placed at a certain time with specific items) can remain as stable, long-term data.
Stables vs Volatiles
Separate stable reference values from values that drive change.
Red Lines at the Architecture Level
Vellamon Kalliot defines red lines directly at the architectural level. These red lines clearly specify what agents and clients are allowed and not allowed to do with data. Because these boundaries are structural rather than purely application-level, they act as preventive controls instead of reactive ones. This is especially critical in the AI agentic era, where autonomous agents can operate at very high speed and may produce unpredictable or incorrect actions.
Why Vellamon Kalliot Matters
Modern data environments face growing complexity due to rapid technological change, increasing automation and AI, and stricter regulatory requirements.
Traditional approaches often attempt to manage this complexity by adding more rules and modeling constructs. Vellamon Kalliot takes a different approach by focusing on reductive clarity — creating simple but strong structural boundaries that protect what must remain stable while allowing other parts of the system to change.
Vellamon Kalliot is particularly valuable because it:
- Reduces cognitive complexity through a simple, memorable pattern.
- Supports long-term data durability independent of specific technologies.
- Makes it easier to maintain accountability and traceability in multi-agent environments.
- Defines preventive red lines at the architectural level, which is essential in the AI agentic era.
- Helps avoid building rigid systems that become expensive and risky to change.
- Provides a practical way to handle regulatory requirements such as GDPR by clearly separating erasable information from long-term business records.
Summary
Vellamon Kalliot is a thinking model that helps data architects and organizations approach data architecture with clarity and long-term perspective. By using the metaphor of water and cliffs, and by emphasizing clear structural boundaries (including the distinction between valuable Data and erasable Logs, as well as preventive red lines), it provides a practical way to design systems that can withstand rapid change while maintaining integrity, accountability, and understandability.
It completely removes the need to define or debate business models and semantic meanings. Instead, it focuses on determining what data must be protected as immutable and at which stage. This makes the model especially effective in complex, fast-changing, and AI-driven environments.
Vellamon Kalliot is fundamentally about the inherent physics of data — its natural behavior, properties, and the structural forces needed to manage it reliably over time — rather than any technical tricks, patterns, or implementation details.
Vellamon Kalliot does not replace existing modeling techniques but serves as a foundational thinking layer that helps ensure those techniques are applied with the right priorities — protecting what must remain stable while enabling the flexibility that modern systems require.
