endorsedCompleted

Redshift Data Warehouse & ML Models for System Loco

Services Covered:

AI and Machine Learning

Industry

Information Technology

Duration

6 months

Budget

Confidential

Start Date

1 August 2024

Team Size

1-5 Employees

Engagement Model

Offshore

Project Capability Score (PCS)

82/100

Strong

Project Capability Score (PCS) estimates how capable Simform is of successfully delivering a project like this, based on its past experience, track record, and ability to handle similar work. Learn more about PCS

Client

System Loco

Nottingham, United Kingdom

Project Summary

Simform built a Redshift-based data warehouse and developed machine learning models for System Loco, a shipment monitoring company. The engagement included ETL pipelines from MongoDB to S3/Redshift and predictive models for battery life estimation and zone-based movement, resulting in improved ETL reliability and optimized storage utilization.

Key Challenges

Build a data warehouse using AWS RedShift
Develop ML models to provide feedback on System Loco devices and data

Project Deliverables

Redshift-based data warehouse (Redshift cluster setup)
ETL pipelines from MongoDB to Amazon S3 and Amazon RedShift
Data schema and indexing strategies for RedShift
Documentation of data flow, transformations, and RedShift setup
Battery life estimation ML model with API/service deployment
Zone creation and movement prediction ML model with API/dashboard

Project Solution

Simform created a centralized Redshift data warehouse by extracting MongoDB collections to Amazon S3 and implementing ETL pipelines to load and transform data into Amazon RedShift. They designed schemas and indexing strategies to optimize query performance and ensured data consistency. In parallel, the team developed machine learning models for battery life estimation and zone-based movement prediction, validated models, and delivered them as APIs or services with technical documentation.

Project Outcome

Improved ETL pipeline reliability
Reduction in query response time
Optimized storage utilization
Improved data availability and accuracy
Battery life estimation model developed, validated, and deployed
Zone identification and movement-prediction models implemented with provided APIs/dashboards

Platforms

Tech Stack

Amazon RedShift
MongoDB
Amazon Web Services (AWS)

Client Endorsement

Overall Review Rating

4.75

Timeliness

Cost Rating

Willing to Refer

Quality of Deliverables

“Simform's work received positive feedback from the client, optimized storage utilization, and improved ETL pipeline reliability. The team was proactive, responsive, and adaptable and communicated regularly. Moreover, the team's structured approach and commitment to high-quality results stood out.”
Kourosh Amouzgar
Data Scientist
Date: 03/18/2025