Shardul Manohar

AI/ML Engineer with 3+ years of experience building, training, and deploying machine learning and deep learning systems across industrial and research environments. Strong hands-on expertise in computer vision, predictive modeling, vector embeddings, and LLM-based systems, with experience integrating models into production workflows. Comfortable working across the full ML lifecycle—from data preparation and experimentation to deployment, monitoring, and iteration.

AI/ML Engineer at Optra Scan Pvt Ltd (2025-07 – Present)

Built and trained deep learning models for disease diagnosis from cytology slides and Tuberculous bacilli detection in medical imaging.

• Built and trained deep learning models for disease diagnosis from cytology slides, and Tuberculous bacilli detection.
• Applied CNN-based architectures (YOLO, EfficientNet, custom Keras models) for medical image analysis.
• Developed end-to-end computer vision pipelines for digital pathology, from preprocessing to inference.
• Collaborated with domain experts to refine data labeling, evaluation metrics, and model performance.

AI/ML Engineer at Demos Foundation centre for Research in Humanities (2025-02 – 2025-07)

Designed and implemented vector embedding pipelines, similarity search systems, and RAG-based document understanding solutions.

• Designed and implemented vector embedding pipelines for large-scale document and image datasets.
• Built similarity search and duplicate detection systems using FAISS.
• Developed RAG-based systems for document understanding and semantic retrieval.
• Implemented OCR, NER, and sentiment analysis pipelines using spaCy and related NLP tooling.

Software Engineer at AutomationEdge (2022-11 – 2025-01)

Developed ML-powered components within an enterprise automation platform, including deep learning pipelines and clustering algorithms.

• Developed ML-powered components within an enterprise automation platform.
• Designed a deep learning pipeline for image-based table extraction from documents.
• Implemented a novel Best-K selection algorithm for K-Means clustering, improving clustering stability.
• Integrated ML models into production RPA workflows via APIs and automation logic.
• Contributed to GenAI initiatives, including LLM-assisted code generation and model refinement.

BS–MS in Dual Degree in Science – IISER-Pune (2023)