MADHUMATHI RAJESH

Cost Efficiency Enhancement:

• Built and spearheaded development of Genie, a comprehensive package comprising 15 use cases encompassing various summarization tasks, information extraction tasks, and chatbots using Gemini Pro model in Vertex AI. Genie is exposed as API to operations, facilitating various Proof of Concepts (PoCs).Strategically transitioned from Google ASR engine to Whisper AI in Vertex AI, reducing daily audio processing costs by 78.33%.
• Achieved 4x reduction in costs, translating to tangible impact on organization's financial performance.

Code Optimization and Workflow Enhancement:

• Led code optimization strategies, reducing model prediction time from 0.38s to 0.07s per document, enhancing operational efficiency.
• Established robust governance frameworks for data and modeling, ensuring compliance and best practices, and fostered effective communication with Chief Data Officer (CDO).

Responsible AI Framework Development:

• Led development of industry-leading Responsible AI framework, covering governance, EU AI regulations, GDPR compliance, and collaborating with Stanford University scholars.

Productiozation and Recognition:

• Transformed framework into a valuable product, resulting in a 30% increase in product adoption.

Client Acquisition and Project Wins:

• Leveraged Responsible AI framework to secure high-profile client projects, contributing to a 20-25% increase in client portfolio.

Proposal Management and Project Delivery:

• Instrumental in preparing RFP responses, leading to a 40 - 45% increase in conversion rate of RFPs to successful project acquisitions.

Financial Impact:

• Secured projects with a remarkable income, resulting in a 32% increase in revenue attributed to Responsible AI initiatives.

Cheque Clearing Project:

• Spearheaded project, saving 3 FTE per country and streamlining communication processes.
• Implemented advanced technologies: VGG16 for data annotation, Faster R-CNN for object detection, and OCR with LSTM for manual intervention reduction.

ARA: Advanced Risk Analytics:

• Pioneered anomaly detection using cutting-edge algorithms, resulting in a 20% reduction in fraud-related incidents and substantial financial savings.

Trade Automation:

• Implemented streamlined trade automation system utilizing custom NER models and feature engineering integrated with RPA.

Skynet:

• Executed Skynet project, achieving efficient multi-label text classification using SS3 algorithm and USE embedding with sequential nets.

MT599 Information Extraction and Message Routing:

• Spearheaded MT599 Message Routing project, utilizing SVC and Logistic Regression for classification.
• Deployed solution on AWS for continuous model monitoring and sustained accuracy.

ISO Rejection Message Prediction:

• Implemented classification model using Logistic Regression, recognized as first ML model deployed in production with 98% accuracy.

• Performed system analysis, documentation, testing, implementation, and user support for platform transitions. Validated results and performed quality assurance to assess accuracy of data. Assessed data modeling and statistics to integrate high-level business processes with data rules. Identified and resolved problems through root cause analysis and research.
• QC inline Automation: Built table detection algorithm (TableNet) and Entity extraction model using custom object detection algorithm, resulting in 6 FTE Save.
• Image Enhancement for Kofax documents: Image processing approaches and reverse engineering-based Auto encoder for Water mark and Salt & Pepper noise removal increased document processing speed to 3X.
• Loan Queue Allocation for Paycheck Protection Program (CARES Act): Classified loan applications based on document availability in repository, time stamp, and additional natural language comments using XGBOOST and Random Forest classification models.

• Mastering Text handling techniques to construct solutions for various Natural Language applications.
• Been part of 2 funded projects.
• Reviewed program materials and coordinated updates to keep department materials relevant and accurate.
• Designed and implemented various educational activities and programs to meet department needs.

• Collaborated with colleagues on curriculum revision, evaluation of course syllabi and lesson plans for AI curriculum.
• Evaluated and supervised student activities and performance levels to provide reports on academic progress.
• Applied innovative teaching methods to encourage student learning objectives.
• Established research agenda AIML field to publish in peer-reviewed journals.
• Facilitated academic and community collaborations to increase number of community-engaged research proposal submissions to extramural funders.
• Revised course objectives, course materials, instructional and assessment strategies elective courses.
• Performed research to serve as basis for academic writing for publication.

• Arranged nearly 48 Seminars, 8 Workshops (On board and Outside Campus)
• Had also been Class-Advisor, Timetable in-charge and Project Coordinator.
• Evaluated and supervised student activities and performance levels to provide reports on academic progress.
• Created materials and exercises to illustrate application of course concepts.
• Applied innovative teaching methods to encourage student learning objectives.
• Mentored students and communicated internship and employment opportunities.
• Proctored exams and provided remediation for learning improvement goals.