Project overview
- Organisation
- Safehouse Tech
- Role
- Lead Data Analyst
- Period
- Sep 2022 - Jul 2023
Real-time Cybersecurity Event Intelligence
Near-real-time security-event pipelines on Google Cloud with anomaly detection, automated alerting and Generative AI-assisted threat classification, improving analytical throughput by approximately 25% and contributing to an approximately 15% reduction in system downtime.
~25%improvement in analytical throughput
Context
A cybersecurity product business generating high-volume security events that needed to be analysed close to real time to detect anomalies and respond to threats.
My Role
As Lead Data Analyst I led a 7-person analytics and engineering team and partnered directly with the CPO and engineering leadership on security intelligence, anomaly detection and real-time analytical capabilities.
Approach
- Near-real-time security-event processing on GCP using BigQuery, Dataflow and Pub/Sub.
- Anomaly detection over event streams with automated alerting.
- Generative AI-assisted threat classification and workflow automation to raise analytical throughput.
Business Impact
- Approximately 25% improvement in analytical throughput.
- Contributed to an approximately 15% reduction in system downtime.
Technology
- BigQuery
- Dataflow
- Pub/Sub
- Google Cloud Platform
Related work
- Burda Luxury, Asia Pacific
AI-enabled Analytics Agent
Conversational analytics built on governed semantic APIs and production workflows.
Read case study - Burda Luxury
Customer Behaviour Forecasting
Customer-behaviour forecasting and audience analytics that improved targeted-advertising precision by approximately 25%.
View project - Burda Luxury
NLP Sentiment Analysis
NLP-based sentiment analysis strengthening understanding of customer and audience behaviour.
View project