Rahul Singh

As an Applied Machine Learning Engineer, I approach complex challenges with a blend of logic and pragmatism, driven by an insatiable curiosity to explore innovative solutions. My determination fuels my ambition to continuously learn and adapt, ensuring I deliver impactful results while remaining flexible in dynamic environments.

Experience

Engineer - Applied Machine Learning

Intangles Lab Pvt. Ltd. · Apr 2023 – Present

Not yet confirmed

• Extended and optimized production fuel analytics algorithms on OBD telemetry data, including DBSCAN + LOESS based event detection for refill, theft, and data-loss alerts.
• Built sub-threshold fuel event detection to capture small thefts/refills missed by existing threshold logic and sensor limitations.
• Built data-loss detection for missing fuel level, GPS, and telemetry signals during active vehicle movement, separating sensor dropout from genuine fuel events.
• Applied SMA-based smoothing to noisy fuel-level sensor data before classification, reducing false positives by 15%.
• Refactored the Node.js fuel analytics codebase, fixing level-consumption mapping and bin-based mean/SD computation, reducing support queries and reclaiming 30+ hours weekly.

I

Machine Learning Engineer

Intangles · Apr 2022 – Present

Not yet confirmed

• Extended and optimized a core DBSCAN + LOESS fuel event detection algorithm on OBD telemetry data, maintaining 95% alert accuracy across refill, theft, and data-loss events at scale.
• Built a sub-threshold fuel event detection system from scratch, capturing small thefts and refills previously missed by sensor limitations and existing threshold logic — now used as a core production feature in fuel analytics workflows.
• Built a data-loss detection system identifying missing fuel level, GPS, and telemetry signals during active vehicle movement, separating sensor dropout from genuine fuel events.
• Applied SMA-based smoothing on noisy fuel-level sensor data to extract the underlying signal before classification, reducing false positives by 15%.
• Refactored the Node.js fuel analytics codebase, fixing bugs in the level-consumption mapping and bin-based mean/SD computation - reclaiming 30+ hours weekly by reducing support queries from 6 to 2 per day.

Junior Machine Learning Engineer

Intangles Lab Pvt. Ltd. · Apr 2022 – Apr 2023

Not yet confirmed

• Worked on production ML and analytics workflows for vehicle telemetry and fuel analytics systems.
• Built and debugged Node.js/Python-based analytics logic, statistical features, and data quality checks for fuel-level and telemetry signals.
• Supported fuel event detection workflows by improving data processing, alert reliability, and analysis of noisy sensor data.