Skip to content
plainsight.pro

Why Technical Guidelines

These Technical Guidelines guide you towards a better implementation within a Data Project at Plainsight. These guidelines result from years of experience, countless projects and experiences. Projects at Plainsight should, as closely as possible, adhere to these best-practices.

While these Technical Guidelines provide our way of working, there are always reasons one prefers to use another way/manner. The flowchart below shows what to do when. 

---
config:
  theme: forest
  layout: elk
---
%%{init: { "flowchart": { "useMaxWidth": true } } }%%
graph TD;
    A[Q? Can I adhere to the technical Guidelines?]
    B[Yes, it's a greenfield.]
    C[Implement the Technical Guidelines.]
    D[No, the technical guidelines are not sufficient.]
    E[Alter the 'Git' Page of the Technical Guidelines and add your new experiences.]
    F[No, I'm working on a project that differs from it.]
    G[Document why you differ from these Technical Guidelines on the as-is scenario. For future scope...]
    

    A --> B --> C
    A --> D --> E --> C
    A --> F --> G --> C
    

    classDef pinkBox fill:#FDCAD2,stroke:#031B89,stroke-width:1px;
    class B,D,E,F,G pinkBox;

    classDef blueBox fill:#031B89,stroke:#031B89,stroke-width:1px,color:#FFFFFF;
    class A,C blueBox;

Why do we have these Technical Guidelines?

  • We want to provide best-practices to our customers.
  • Our projects should easily be transferrable between consultants.

Looking ahead: Technical Guidelines in an AI-first world

These Technical Guidelines are not just written for humans reading a wiki; they are the foundation for how our AI assistants will reason about data projects at Plainsight.

Making the Playbook available to agents

We are actively working on exposing this content to the AI agents our customers use, via a Model Context Protocol (MCP) integration. That means: - Agents can retrieve and reference the latest Plainsight best-practices directly from the Playbook. - Guidance stays centralised, versioned and explainable, instead of being hidden in ad‑hoc prompts. - Changes to the Playbook automatically update the "source of truth" that agents use when helping to design or review solutions.

[!tip] Data Engineers as AI "orchestrators" We expect a future where Data Engineers increasingly steer fleets of AI agents that design, generate and maintain ETL and analytics solutions. In that future: - These Technical Guidelines become machine-consumable instructions, not just reading material. - ETL, modeling and platform decisions are implemented by agents, while humans focus on architecture, validation and exception handling. - Deviations from the guidelines are explicitly documented and can be surfaced by agents during code review or solution design.

This is why clarity, consistency and explicit trade‑offs in our guidelines matter so much: they make it easier for both humans and AI agents to build solutions that look and feel like Plainsight.