Skip to content
All projects
Internal platform · DataOngoing

Internal Operations Platform

Architected a central internal platform with a single data model and role-based workflows to replace forms, spreadsheets and manual approvals — prepared for automation and AI.

Context

Internal processes were spread across forms, emails, spreadsheets, SaaS tools and manual approvals, with no central place to run or audit them.

Challenge

  • Fragmented processes with no single owner or data model.
  • Manual approvals and duplicated tools.
  • No reliable foundation for automation or AI.

My role

Architect and lead engineer for the internal platform and its data model.

Responsibilities

  • Architecture of a new internal platform
  • Central data model design
  • Role-based access
  • Workflow design
  • System integrations
  • Replacement of legacy processes
  • Preparation for AI-assisted functionality
  • Preparation for removal of duplicate tools

Architecture

  • A central data model in PostgreSQL/Supabase as the single source of truth.
  • Role-based access controlling who can see and do what.
  • Workflows and approvals modelled as first-class, auditable objects.
  • Integration points and an API-driven design ready for automation and AI.

Fragmented

  • Fragmented processes with no single owner or data model.
  • Manual approvals and duplicated tools.
  • No reliable foundation for automation or AI.

Connected

  • A central data model in PostgreSQL/Supabase as the single source of truth.
  • Role-based access controlling who can see and do what.
  • Workflows and approvals modelled as first-class, auditable objects.
  • Integration points and an API-driven design ready for automation and AI.

Project lifecycle

  1. 01

    Fragmented tools

    Forms, email, spreadsheets, manual actions.

  2. 02

    Architecture

    Central data model and platform designed.

  3. 03

    Workflows

    Role-based, automated workflows.

  4. 04

    Integration

    Identity and systems connected.

  5. 05

    Reporting

    Reporting on a single data model.

  6. 06

    AI-assisted

    Prepared for AI-supported operations.

Implementation

  • A modern frontend over a Supabase/PostgreSQL backend.
  • REST APIs connecting the platform to identity and operational systems.
  • Automation replacing manual, repetitive steps.
  • AI-assisted functionality introduced where it adds value, with controls.

Documentation & handover

Operational and project documentation were prepared as part of delivery, with handover and acceptance support included where in scope — so the environment can be operated, audited and improved after go-live.

Client names and selected implementation details are intentionally anonymized.

Outcome

A central foundation for internal operations, automation and future AI-supported workflows.

Related projects