Skip to main content
Case studies

How we take a requirement through to a working system.

Each case study states the problem, the architecture, the security posture and the honest current status. We do not publish client names, revenue figures or performance metrics we cannot evidence.

Portfolio

Engineering case studies

LiveEnterprise systems

BMMIS — business management & monitoring information system

Internal OMS IT™ enterprise system

Problem

Operations, HR, finance and reporting were tracked in separate places, so owners had no single, current view across ventures.

Approach
  • Consolidate the operating record into one system of record with clear module boundaries.
  • Model roles first, so reporting and data access follow the organization chart.
  • Ship module by module rather than as a single large release.
Architecture
  • Modular application layer over a relational system of record.
  • Role-based access enforced server-side at the data layer.
  • Reporting views derived from the same source of truth as operational screens.
Technology
  • Web application
  • Relational database
  • Role-based access control
  • Reporting layer
Security
  • Server-side authorization for every data path.
  • Least-privilege roles per module.
  • Change history retained for operational records.
Implementation
  • Module-by-module rollout with the operating team using each module before the next was built.
  • Reporting added after the underlying records were stable.
Current status

Live and in continuous use, with modular expansion and an AI insights layer planned.

Evidence note

Status reflects OMS IT™ internal use of the system. No external client metrics are claimed.

In developmentMedia & publishing

Desher Alo — AI media operating system

Desher Alo Online (OMS ecosystem media venture)

Problem

Editorial work, distribution and audience insight were handled through disconnected tools, making it hard to keep a consistent publishing workflow.

Approach
  • Map the newsroom workflow end to end before introducing any AI assistance.
  • Use AI to assist editorial steps — drafting support, verification prompts, summarisation — with journalists retaining the decision.
  • Keep publishing authority with named human editors.
Architecture
  • Editorial workflow application with staged review.
  • Assistive AI layer invoked per editorial step, not as an autonomous publisher.
  • Distribution and audience data reported back into the same workflow.
Technology
  • Web application
  • Content workflow model
  • LLM-assisted editorial tooling
Security
  • Editorial roles and publishing permissions separated.
  • Human approval required before publication.
Implementation
  • Newsroom pilot first, with a narrow set of assisted steps.
  • Wider syndication and creator tooling deferred until the pilot workflow is proven.
Current status

In development — newsroom pilot stage.

Evidence note

Status is self-reported project maturity. No audience, revenue or performance figures are claimed.

See the full venture list on Projects and the technical model on System Architecture. Our delivery process is documented on How We Build.

Chat with OMS on WhatsApp