AI/ML Engineer Internship in Bangalore at Gray Mobility

We are looking for an AI/ML intern who can build the intelligence layer for Gray's Smart Ambulance platform. The Ambulance Central Module (ACM) and medical device integration will be supported by suppliers and third-party integrators. Your primary focus will be on building AI/ML models, analytics, alerting logic, and decision-support features on top of the data streams (edge + cloud).

Selected intern's day-to-day responsibilities include:

1. Analyze real-time vital signals by detecting anomalies in SpO2, heart rate, and blood pressure while identifying trends and early warning signals
2. Design clinical decision support primitives including triage scoring, risk stratification, and alert prioritization
3. Detect events from sensor streams such as patient movement, fall risk, AED usage, and cabin context signals
4. Predict emergency response ETAs and optimize route intelligence to reduce response times
5. Trigger green corridor intelligence by designing traffic priority logic and signal orchestration events
6. Build analytics solutions including cohort dashboards, response quality KPIs, outcomes analysis, and incident reports
7. Develop continuous learning pipelines through dataset curation, labeling, evaluation, and retraining strategies
8. Own the full machine learning lifecycle from problem framing and data preparation to feature engineering, model training, evaluation, and deployment
9. Build machine learning pipelines for time-series data including medical vitals and vehicle telemetry
10. Develop APIs and microservices to expose machine learning outputs such as alerts, scores, and predictions to dashboards and applications
11. Implement explainability and confidence reporting to justify why critical alerts are triggered
12. Create experiment tracking systems, model versioning processes, and reproducible training scripts
13. Collaborate with suppliers and integrators to handle device data formats and consume normalized streams from ACM or edge gateways
14. Deploy machine learning services on AWS using containerized or serverless architectures and configure monitoring for drift, latency, and accuracy
15. Document datasets, assumptions, and model limitations clearly to support safety-critical usage

Key ML problem areas:

1. Time-series anomaly detection (unsupervised + supervised)
2. Classification/regression for risk scoring and outcome prediction
3. Sequence models (LSTM/GRU/Transformers) for multi-vital streams (optional)
4. Signal processing/smoothing/artifact detection (noise removal from vitals)
5. Computer vision (optional): cabin monitoring/patient posture/safety events

Key features to support - Roadmap:

1. Biometric patient identification
2. Real-time multi-vital monitoring
3. Big data analytics
4. IoT-enabled medical devices
5. Cloud-based healthcare systems
6. Mobile telemedicine
7. Automatic external defibrillator (AED)
8. GPS and navigation system
9. Environmental sensors
10. Data transmission protocols
11. Integration hub
12. Data security & privacy features
13. Patient vital information display inside ambulance
14. Vehicle edge computing
15. Cabin patient monitoring system (CPMS)
16. Dispatch & call center platform (This feature is majorly for drivers)
17. Hospital integration platform
18. Telemedicine platform
19. Driver/Co-driver UX system (DCUX)
20. Compliance & security platform
21. Drone integration system
22. Smart ambulance traffic priority system (green corridor) - traffic server integration with nearby signals

Success metrics/deliverables:

1. One deployed ML service producing alerts or scores from real-time data
2. Demonstrated reduction in false alerts via artifact detection and threshold tuning
3. Dashboards or reports summarizing alert performance (precision and recall) and response KPIs
4. Green corridor logic prototype integrated with route and signal events
5. Reproducible training and evaluation pipeline with documented assumptions

1