End-to-end data platform powering a bespoke analytics app
I was brought in to build a greenfield data platform and the infrastructure around it. I designed and delivered the full stack, ingestion, modelling, orchestration, and a secure API layer that serves data to both the bespoke client‑facing web app and external consumers.
About Rubik
Rubik helps ambitious manufacturing, construction, and engineering companies harness digital tools to scale smarter, move faster, and stay ahead. Their work spans digital consulting, data & AI services, and finance transformation, combining strategy with hands‑on delivery.
Learn more on their website: rubiktech.co
The challenge
Design a standardised platform that can be deployed across clients. It needs to support multiple ingestion connectors, keep each client’s data stored separately, enforce strong security, and remain cost efficient to run and scale.
Key needs
- Standardised, reusable data platform that can be rolled out across clients
- Support for varied data ingestion connectors (APIs, files, databases)
- Strict data isolation per client with separate storage
- Robust, versioned API layer to serve data to the app and external clients
- Cost‑efficient architecture with a clear scaling path and minimal ops overhead
Approach
I began with discovery and scope, then ran collaborative design sessions with business stakeholders to validate the architecture. Once agreed, I issued a detailed plan with defined days for full cost clarity and delivered the platform end‑to‑end.
What I did
- I began with discovery and scope, then ran design sessions with stakeholders to validate architecture
- I implemented all cloud infrastructure, IaC, ingestion pipelines, modelling, orchestration, and monitoring
- I designed and built a secure, versioned API layer that serves curated datasets to the app and to clients
- I set up multiple environments and CI/CD for safe development and repeatable releases
- I defined data contracts with the frontend and wired the app to the data API
- I documented everything and provided knowledge transfer for sustainable ownership
Outcomes
The result is a fully functioning analytics web app fed by the new data platform through a secure, versioned API. By handling the full stack myself, the client avoided needing to engage multiple engineers while keeping costs predictable and low. They now have solid foundations to scale their platform and start exploring AI, ML, and agentic capabilities.
Results at a glance
- Live analytics web app powered via a secure, versioned API on top of the new data platform
- End‑to‑end build delivered by a single engineer, reducing time‑to‑value and dependency on multiple teams
- Low, predictable run costs with autoscaling, monitoring, and clean handover docs/runbooks
- Clear path to ML, predictions, and agent workflows thanks to well‑modelled data and stable API contracts