Selected Work

Systems in production.

A selection of recent engagements across software, AI, data and cloud. Each one started with a complex problem and ended with a system the client's team runs every day.

Operations Platform

A unified internal platform that replaced spreadsheets and six disconnected tools with one system for dispatch, inventory and reporting.

Industry: Logistics
Scope: Architecture, platform build, rollout
Stack: TypeScript, Node.js, PostgreSQL

Challenge

Operations ran on shared spreadsheets and manual hand-offs. Data was re-entered several times a day and no one had a reliable view of the current state.

Approach

We modelled the core workflows first, then built a modular platform with a typed API, role-based access and integrations to the existing accounting system.

Outcome

One source of truth for daily operations, far less manual entry and a codebase the internal team now extends on its own.

Document Intelligence

An AI pipeline that classifies and extracts information from high-volume documents, with human review built into the workflow.

Industry: Financial Services
Scope: AI pipeline, review tooling, integration
Stack: Python, FastAPI, LLMs, RAG

Challenge

Thousands of documents a week were read and keyed in by hand. Turnaround was slow and errors surfaced late in the process.

Approach

We combined document classification with retrieval-augmented extraction, confidence scoring and a review queue, so people only handle the cases the model is unsure about.

Outcome

Faster processing, consistent structured data and a full audit trail of every automated and human decision.

Unified Data Platform

A modern data stack that consolidates operational sources into a single, trusted analytics layer.

Industry: Retail
Scope: Pipelines, warehouse, data models
Stack: Airflow, dbt, BigQuery

Challenge

Sales, inventory and e-commerce data lived in separate systems. Every report meant a manual export, and teams disagreed on basic numbers.

Approach

We built orchestrated ingestion pipelines, a layered warehouse and tested, documented data models with shared metric definitions.

Outcome

Daily refreshed dashboards from one trusted source and a foundation ready for forecasting and machine learning.

Cloud Modernization

Migration of legacy workloads to cloud infrastructure defined as code, with observability and cost controls from day one.

Industry: Healthcare
Scope: Cloud architecture, migration, DevOps
Stack: Azure, Terraform, Docker, CI/CD

Challenge

Critical applications ran on ageing on-premise servers with manual deployments, limited monitoring and growing maintenance costs.

Approach

We containerised the workloads, defined every environment in Terraform and moved services across in stages behind automated pipelines.

Outcome

Repeatable, audited deployments, clear visibility into system health and infrastructure spend that is tracked and controlled.

Scheduling SaaS Platform

A multi-tenant SaaS product for service businesses, taken from validated prototype to a paid, production platform.

Industry: Professional Services
Scope: Product engineering, billing, analytics
Stack: ASP.NET Core, React, PostgreSQL, Azure

Challenge

The founders had a working prototype and early customers, but no architecture for tenancy, subscriptions or scale.

Approach

We re-architected the product for multi-tenancy, added subscription billing, usage analytics and an automated release process.

Outcome

A stable platform onboarding new customers without engineering involvement, and a roadmap the team can ship against.

Client names are withheld at their request. Detailed case studies are available on request.

Have a system that needs engineering?

Tell us about the problem. We will reply within two business days.

Start a Conversation