Looking for a job

Sumitra Shukla

@sumitrashukla

Machine Learning Engineer at IntelliRAG Pvt. Ltd., where I developed healthcare AI, a clinical notetaker, and a RAG system for homeopathy literature.

India

Message

At IntelliRAG Pvt. Ltd., I developed AI-enabled healthcare solutions, including a smart healthcare management system for patient and doctor profiles, appointments, availability, and vital records.

I also developed a WebSocket-enabled clinical notetaker that used speech recognition to capture doctor–patient conversations and support clinical-note and prescription generation. For homeopathy literature, I implemented a Retrieval-Augmented Generation system and fine-tuned a T5-based small language model on synthetic and curated datasets.

At IIT(ISM) Dhanbad, my postdoctoral research focused on phonocardiogram signal analysis for heart valve disease detection. I developed a heart sound segmentation technique using kurtosis and zero-frequency resonator features with HMM-based modeling.

My research also covers stressed-speech recognition and rotating-machinery fault diagnosis using time-series vibration data. I hold a Ph.D. from Indian Institute of Technology Guwahati, where my thesis examined spectral analysis of stressed speech for speech recognition.

Experience

Work history, roles, and key accomplishments

Machine Learning Engineer

IntelliRAG Pvt. Ltd.

Nov 2023 - Mar 2026 (2 years 4 months)

Managed a cross-functional team of eight members, providing technical direction and monitoring progress of AI-enabled healthcare solutions. Designed an AI-enabled smart healthcare management system and developed a WebSocket-enabled clinical notetaker with speech recognition.

PostgreSQL DynamoDB S3 WebSockets Speech Recognition RAG T5 Transformer Fine Tuning Machine Learning Deep Learning

Postdoctoral Researcher

IIT(ISM) Dhanbad

Jan 2017 - Dec 2019 (2 years 11 months)

Developed a heart sound segmentation technique using kurtosis and zero-frequency resonator features, combined with HMM-based modeling for cardiac sound duration analysis.

Signal Processing HMM Kurtosis Zero Frequency Resonator Heart Sound Analysis

Research Fellow

IIT Guwahati

Jan 2006 - Dec 2007 (1 year 11 months)

Investigated isolated-word recognition in spontaneous speech using speech signal processing. Extracted MFCCs as acoustic features and applied Vector Quantization for word-level modeling and recognition.

Speech Processing MFCC (Feature Extraction)Model Quantization Keyword Spotting

Research Collaborator

IIT (ISM) Dhanbad

Collaborating on research in machine learning and deep-learning-based fault diagnosis of rotating machinery using time-series vibration data. Developing signal-processing and AI/ML methods for bearing-condition monitoring and cross-dataset fault classification.

Machine Learning Deep Learning Signal Processing Fault Diagnosis Vibration Analysis

Education

Degrees, certifications, and relevant coursework

Indian Institute of Technology (ISM) Dhanbad

Postdoctoral Researcher, Biomedical Signal Processing

2017 - 2019

Conducted postdoctoral research on phonocardiogram signal analysis for heart valve disease detection, developing heart sound segmentation techniques using kurtosis and zero-frequency resonator features.

Indian Institute of Technology Guwahati

Ph.D., Speech Signal Processing

2007 - 2014

Pursued a Ph.D. in spectral analysis of stressed speech for speech recognition, developing an HMM-based recognition system for emotional and stressed speech conditions.

Indian Institute of Technology Guwahati

Research Fellow, Speech Signal Processing

2006 - 2007

Worked as a research fellow on keyword spotting in continuous speech, extracting MFCC features and applying vector quantization for word-level modeling.