Research Projects

🧠 LangChain Anomaly Detection Agents

Developed intelligent agents using LangChain for anomaly detection in logs and data streams, showcasing the power of LLMs in monitoring systems.
🔗 GitHub: LangChain Anomaly Detection Agents

📚 Research Assistant Agent

Designed an AI agent using AutoGen to assist in literature review, summarization, and research automation tasks.

• Tech Stack:

  • AutoGen: Framework for building LLM-powered agents and tools

  • Tavily Python SDK: For semantic search and retrieval

  • Groq Python SDK: Interacting with Groq’s language models

  • Pydantic: Data validation with type annotations

  • python-dotenv: For managing environment variables securely
    
    

🔗 GitHub: Research Assistant Agent 

📊 Deck Automation Tool

Created an open-source solution for automating presentation deck creation using Python and templating.
🔗 GitHub: Deck Automation

🔍 Electronic Design Automation (EDA) Chatbot with LLM Integration

Developed an AI-powered streamlit based chatbot using internal XML-based knowledge to assist engineers in creating high-level verification plans through intent-based queries. Integrated LLMs and APIs for seamless usability across design projects. Deployed internally at Synopsys.

⚙️ Automated Verification Requirement Extraction

Designed an ML-based solution to extract RTL verification needs from complex PDFs by analyzing layouts, identifying text, tables, and diagrams. US Patent granted (2022, Synopsys).

🧾 Register Attribute Automation

Built an AI-driven tool to extract register attributes from specs and auto-generate output in Excel, enabling code generation in C, SV, and RTL. Integrated with verification workflows. US Patent filed (2024, Synopsys).

🤖 AI Innovation in EDA

Pioneered AI techniques for extracting design requirements and generating test cases in EDA. US patent filed (2025, Synopsys).

💡 LLM-Based Hardware Verification

Introduced a novel method for hardware verification using natural language and LLMs, significantly reducing verification time. Published at DVCON India 2024.

Achievement:- She has demonstrated exceptional leadership in promoting AI technologies within EDA, receiving recognition through the Q2 2023 SDG award for her team leadership, research publications, and patent contributions. 

📈 ML-Based Engineer Efficiency Tracking

Built an ML-driven solution using Perforce and Jira logs to analyze developer activity and send automated performance reports.
Deployed internally at Synopsys.

🎥 Computer Vision for Surveillance & Safety

Contributed to video analytics solutions for security systems including explosive detection, no-helmet alerts, crowd density tracking (e.g., Maha Kumbh), and COVID hygiene compliance.
Techniques used:

• Object detection with YOLO & SSD (fine-tuned for crowded scenes)

• Multi-object tracking using Deep SORT & multi-hypothesis frameworks

• Human pose estimation for region-of-interest detection

• Image classification & clustering using SIFT, HOG, Autoencoders, etc.

• Preprocessing with denoising, Hough transforms, and super-resolution

• Video summarization and anomaly detection

• Clustering of under-vehicle images

Deployed internally at Vehant Technology Pvt. Ltd.

🧬 COVID-19 Forecasting with O-ARIMA

Proposed the O-ARIMA model to forecast COVID-19 spread with improved accuracy.
Published in Applied Soft Computing (2022).

🧪 COVID-19 Severity Prediction

Identified key biomarkers and developed an interpretable AI model for risk stratification.
🔗 Online tool available here
First Indian patient cohort study; published in Cureus 2024.

🧠 Neuro Disease Diagnosis Classifier

Developed models to detect schizophrenia and autism using brain connectivity patterns and multivariate signal processing.
Published in Medical Image Analysis (2017, 2019).

🧫 Cancer Risk Group & Survival Prediction

Built deep learning models to predict survival and stage risk in multiple myeloma and breast cancer using Indian cohort data.
Techniques: Decision trees, SVM, XGBoost, clustering (Birch, GMM, agglomerative), elastic net
Published in Computers in Biology & Medicine (2022).

📊 Other Biomedical Projects

• Deep learning for sleep apnea detection from ECG data

• COVID-19 imaging classification survey

• Autism classification & research mentorship

Publications: Cureus 2024, Applied Soft Computing 2022, and more