DDRTP — Dynamic Data Recovery Time Prediction Algorithm

Own from Salesforce (formerly Own Company) — London, UK

The challenge

Static RTO commitments don’t hold up in practice — actual recovery time depends on detection speed, data complexity, and damage scale, all of which change as an incident unfolds, and a fixed number can’t track that.

What I built

Developed an AI-driven algorithm that dynamically predicts data recovery time based on multiple weighted factors — detection time, data complexity, damage scale, and system performance metrics. The algorithm recalibrates predictions in real time as conditions change and alerts users when predicted recovery time exceeds policy-defined RTOs. Implemented within a SaaS application and now the subject of a pending US patent.

Outcomes

• US patent application pending
• Real-time RTO alignment for Salesforce, IaaS, and on-premises environments
• Enabled proactive risk management rather than reactive recovery
• Validated across multiple enterprise client environments

Tech & approach

Probabilistic modelling · SaaS integration · Real-time recalibration