Ayush Srivastava

About

Machine learning engineer focused on trustworthy AI for high-stakes environments, with work spanning computer vision, explainable AI, Bayesian reasoning, and decision intelligence. Experience includes speech AI, geospatial intelligence, and mission-critical research projects.

Experience

• AI/ML Intern
  Abhinnovation Technologies Pvt. Ltd.
  Jun 2026 – Present

  Training and evaluating speech AI models for AI-assisted examination and assessment platforms using large-scale voice datasets; developing speech data preprocessing pipelines and training deep learning models with PyTorch.

Education

• 2028
  Bachelor of Technology (B.Tech)

  Inderprastha Engineering College

  Information Technology · Aug 2024 – Jul 2028

  Information Technology · Aug 2024 – Jul 2028

Skills

Tools / apps / platforms

Projects

• LifeGraph OS
  Trustworthy AI

  Trustworthy digital-twin framework for longitudinal behavioural intelligence.

• QuantaSeis AI
  Multi-modal learning, uncertainty-aware AI

  Disaster intelligence integrating seismic, satellite, and environmental observations.

• BharatVigil
  Decision Intelligence

  Evidence-centric decision intelligence framework for mission-critical operations.

• VyomNetra- Evidence-Based AI Verification Framework for Space Situational Awareness
  Trustworthy AI, Explainable AI, Bayesian Inference, AI Verification, Decision Intelligence, Space Situational Awareness

  Designed a multi-stage verification framework integrating Bayesian evidence fusion, probabilistic reasoning, and confidence-aware assessment for Resident Space Object verification. Proposed a falsification-first verification workflow for mission-critical environments.

• TerraSight - AI-Powered Geospatial Intelligence Platform for Automated Land Encroachment Detection
  Computer Vision, Remote Sensing, Geospatial Intelligence, Trustworthy AI

  Designed an end-to-end geospatial intelligence pipeline for automated land encroachment detection using transformer-based computer vision models trained on satellite imagery. Investigated Swin Transformer architectures for reliable geospatial monitoring.

⚽ Extracurricular activities

• Project Lead — VyomNetra (iDEX Open Challenge)

  Leading the design and development of an explainable AI verification framework for Space Situational Awareness, focusing on Bayesian evidence fusion, confidence-aware reasoning, and trustworthy decision support.

• Team Lead — ClimateCode (Smart India Hackathon)

  Led the development of an AI-powered blue carbon monitoring platform integrating satellite imagery, computer vision, and geospatial intelligence for environmental monitoring and carbon ecosystem assessment.