Brijesh Kumar Karna

• Member of the GenAI Core Team, focusing on the development of Responsible AI frameworks and tools for company-wide applications
• Designed and implemented a Bias Detection model with integrated mitigation features specifically for large language models (LLMs)
• Led the LLM Fairness Evaluation Project to assess and enhance fairness metrics within model operations
• Developed a Privacy Scoring Model to identify and manage user-sensitive information in real-time LLM environments
• Advanced the implementation of Retrieval-Augmented Generation (RAG) methods to align compliance scoring algorithms with the latest AI principles
• Created a Cyber-attack Forecasting Model utilizing the Prophet forecasting tool for daily, monthly, and yearly threat predictions.

2nd Core Team Member, Reporting to VP, Dr. Pethru Raj

• Leadership & Team Management: Hired, handled, and led a talented team of 5 AI Researchers and 8 Research Interns, focusing on multi-domain, multi-vertical AI-related projects
• AI/ML Architecture & POC Creation: Designed and implemented AI/ML architectures and proof-of-concept for different Reliance Jio verticals, including CloudXP, Jio Security, Jio Eva, and Ajio
• JioLLM Development: Received accolades for creating Jio's first JioLLM (similar to ChatGPT) for both internal and external use cases, recognized by Anish Shah, President, COO of JPL
• Virtual Try-On (VTO) Model: Praised by Anand Thakur, CTO, Ajio, for successfully demonstrating a VTO model that enhanced customer experience and reduced returns, impacting 40cr of daily sales
• Call Center Automation: Created a PaaS architecture for call center automation using LLM, supporting popular regional languages
• Innovative Solutions for Various Teams: Developed a Log Analyzer/Anomaly detection POC for the CloudXP team
• Created a POC for an AI-based intrusion detection system for the Jio Infosec team
• Built and trained an end-to-end MLOps platform for the Edge AI Team
• Distributed Edge Computing: Worked on distributed edge computing using Kubernetes (k3s) clustering mechanism on edge devices.

Core Team Member, Reporting to Head of Innovation Lab, Mr Shyam Seshadri

• Building Template Agnostic OCR capabilities for various insured companies like TATA, RELIANCE, SBI, ACKO, GO-DIGIT, UNITED, DHFL, ORIENTAL & THE NEW INDIA
• Building capabilities to onboarding new document types and maintained more than 95% accuracy
• Developed an internal tool for OCR Validation, Which is used by our team to validate documents and generate a review report out of it
• Build a POC for vehicle part detection project, an android app to serve our internal vdd api
• Which was used by our internal team to assess various functionality like records, detect parts of the vehicle and save the video instances etc
• Build a POC for email sentimental analysis for the Customer support team
• Implemented end-to-end computer vision model in AWS sagemaker pipeline for vehicle part detection, and achieved an average 0.4-0.8 milisecond API response time which was previously 6-8 seconds
• Build pipeline for fraud detection, for company wide document/image fraudulent activity tracking
• Implemented multithreading/multiprocessing to build an automation script for parallel claims downloading, monitoring and storing.

Core Team Member, Reporting to CEO, Mr Rajeev Kanth

• Mentoring students and professionals on ML/NLP/AI projects and helping them to understand and stay motivated in this field
• Consulting tech-companies on the implementation of A.I and Machine Learning.

• Team player who can also be independent, prioritize work and thrives in a fast-paced dynamic environment
• Actively participate in challenging software and hardware research projects focused on applying a combination of Deep Learning, Natural Language Processing, and Knowledge Representation pipelines to design, analysis, and engineering workflows for real-world problems.

• Working with the core team, and built the official website of the company from scratch
• Increased website visibility and page ranking from 1,30,000 to 22,000 in 3 weeks around the Globe
• Participated in pre-project analysis and technical assessments to develop a user-friendly interface and correct functionality to meet business objectives.