Back to Success Stories

Transforming a Big Data platform

Fast access to actionable insights from handling 1000x more data

5x Revenue and customer acquisition

>100M EUR In savings generated

1000x More data analysed

Fastest Feature and deployment velocity ever seen at the client

The challenge

Our client, a leading global Spend Optimization Software firm, faced critical roadblocks in evolving their Big Data platform and keeping up with customer demand.

Their system, leveraging AI and Machine Learning for advanced spend data analysis, had a fragile code base that prevented the company from introducing new features and achieving rapid growth through scaling up analysed data volumes.

Thanks to PortBlueSky, our team's productivity and morale have soared. We've successfully navigated tight deadlines and technical challenges, resulting in a stable, high-quality product.
Project Lead (25 devs, 50 overall) - International Spend Optimization Software

Our solution

PortBlueSky designed and implemented a new architecture to break the previous on-premise monolith into smaller parts, which enabled scaling each service on demand.

We carefully lifted each API endpoint one-by-one and delivered a system that was faster & more secure and spearheaded a comprehensive automation revamp to free up developers.

We supplied the majority of the senior and lead positions across frontend, backend, architecture, and DevOps. 

We successfully decomposed the product into manageable, dockerized microservices and seamlessly migrated the database system to PostgreSQL.

Technologies

Backend: Node (TypeScript, fastify, hapi), Python (pyspark, fastapi), Postgres, Redis, Argo, Trino
Frontend: React (TypeScript, next.js, Hooks), Cypress, Storybook, MaterialUI, Grafana, Qlik, PowerBI
AI / Machine Learning: Scikit, PyTorch, Peft
Automation: Kubernetes, Docker, Helm, CircleCI, ArgoCD, Terraform, Splunk, Kibana

Contact

Let’s talk about your challenges and how we can support you.

One of PortBlueSky's tech experts.

Talk directly to a technical director.