Summary

• Experienced Al/ML & Generative Al Solution Architect with an experience of over 13+ years in consulting, architecting, design and development of end-to-end Al solutions to solve complex business problems and optimize processes, thus bridging the gap between business and technology.

• Architect, design and implement generative Al/LLM models and algorithms using state-of-the-art techniques and optimize them for performance, scalability, and efficiency.

• High experience in working with model Fine-Tunning, particularly with open-source large language models like LangChain, Llamalndex and proprietary models like GPT.

• Experience with complex and advanced Retrieval-Augmented Generatiob(RAG) systems, including chunking, indexing, and handling complex PDFS.

• Strong understanding of Prompt Engineering, including the ability of create effective prompts for proprietary models like GPT.

• Computer Vision, Text OCR, NLP, ADvanced Analytics Architect and Cloud Spoc in Al Competency for all industry verticals and Geos.

• Knowledge of common image processing and computer vision software libraries, including OpenCV, Keras, Pytorch, Tensorflow & Tesseract.

• Experience in Mask R-CNN, Detectron2 and Yolo class algorithms for object detection and recognition.

• Automate the end-to-end ML lifecycle with Cloud platforms like Azure, AWS, GCP, Databricks, setup cloud alerts, monitor dashboards, logging and troubleshooting machine learning infrastructure.

• Proven expertise in SQL, NoSQL, MongoDB, DB2, TeraData databases along with Big Data technologies, including Hadoop Ecosystem, Pyspark, Data Lake, and Hive.

• Knowledge on SAS, R, Python, Matlab, Java programming languages for data science applications along with different Al platforms DataRobot, H2O.AI, Spark Beyond, TeraData Al unlimited, Rapidminer & Knime.knimi

• Experienced in setting up CI/CD pipelines to streamline deployment process, leveraging Docker for containerisation and kubernates for orchestration and scaling.

• Currently, working as Generative Al & AI/ML Solution Architect in Deltacubes Pvt. Ltd, leading a team of 12+ Resources.

• Previously, worked as Lead Consultant in Sapiens, Lead Analyst in Gramener, Consultant in Cognizant Associate Business Analyst in Target Corporation, Business Analyst in GenPact.

Technical Skills

AI/ML & Generative Al; LangChain, Llamalndex, GPT, RAG systems (chunking, indexing, complex PDFs), and prompt engineering

Computer Vision & NLP: computer vision, text OCR, NLP, and advanced analytics architect for diverse industry verticals and geographies

Frameworks & Libraries: PyTorch, TensorFlow, Tesseract, Mask R-CNN, Detectron2

Cloud & DevOps: cloud alerts, monitoring dashboards, logging, troubleshooting ML infrastructure, CI/CD pipeline setup, Docker containerization, and Kubernetes orchestration

Databases & Big Data: SQL, NoSQL, MongoDB, DB2, Teradata, Hadoop ecosystem, PySpark, Data Lake, Delta Lake, Hive

Programming Languages: Python, R, Java, MATLAB, SAS

Al Platforms: DataRobot, H2O.ai, Spark Beyond, Teradata Al Unlimited, RapidMiner, KNIME

Work Experience

Timeline: Since August 2023

Role: AI/ML & Generative Al Solution Architect | AI/ML Strategy Leader

• Built and leading the high-performing Research and Development (R&D) team into the nerve center for data-driven insights and AI/ML & Gen Al innovation.

Selected AI/ML Projects:

• Custom Summarization App (Generative Al/LLM/RAG): The app allows users to create custom prompts to summarize PDF files using Al-powered large language models (LLM) like ChatGPT, GPT-4, LangChain, and Llamalndex. Examples: MyMutual FundGPT, Service Assist, Sales Assist.

• Image Synthesis (Generative AI/LLM): Created unique and imaginative visuals based on textual descriptions for Hyper Personalized Marketing Campaigns.

• Voice Assistant (Generative AI/LLM): Developed an Al-powered chatbot in collaboration with Microsoft's SPVoice TTS, enabling multilingual voice command recognition and query responses.

• Fashion Flair (Generative AI/LLM): Built a GAN-based Al-powered virtual fashion stylist/assistant that suggests outfit combinations based on user preferences, fashion trends, and occasions, incorporating TailwindCSS.

• Credit Risk Scorecard (Python/AI/ML): Worked on Application, Behavioural, and Collection Scorecards across various loan products, including Personal Loan, Business Loan, Secured/Unsecured Loan, Loan Against Property, Education Loan, and Home Loan.

• Customer and Marketing Analytics (Python/AI/ML): Developed solutions for Tele-calling Propensity, Top-Up Offers, Cross-Sell, Up-Sell, Lead Scoring, Customer Segmentation, Campaign Analytics, Income Estimation, and Agent Attrition.

• Insurance Analytics (AI/ML): Delivered advanced analytics for Customer Service Scorecards, Lapse Prediction, Surrender Models, Win-back Models, FLS Segmentation, Renewal Models, Recommendation Engines, Customer Lifetime Value (CLTV), Campaign Analytics, and Claim Analytics across Life, Health, Travel, and Auto insurance.

• Mutual Fund/Wealth Management Analytics (Python/AI/ML): Implemented models for Security Analytics, Dividend Forecasting, Market Trend Analysis, Redemption Models, SIP Leakage, Portfolio Analytics, Customer 360, Call Centre Analytics, NPS Analytics, Fund Performance Analytics, SIP Cancellation Prediction, and Salesforce Analytics.

Sapiens India (November 2018 - July 2023) - Bangalore:

• Built Analytics Centre of Excellence across functions, focusing on delivery and advisory capabilities. Created Big Data, Analytics, AI/ML Assets & Capabilities: Developed Agent 360 and Customer 360 views, enterprise data lake, and reusable assets to deliver business impact.

Selected AI/ML Projects:

• Detecting Car Exterior Damage (Deep Learning): Automated detection and quantification of car exterior damages, optimizing property and casualty (P&C) insurance claim processes.

• Intelligent Data Extractor (OCR Tool): Developed an OCR tool to extract data from structured and unstructured formats (Images, PDFs, etc.), supporting dynamic templates and offering multiple output options (APIs, CSV, Database).

• Policy Fraud Model (Classification Model): Built a model for detecting financial fraud in insurance policies, providing early warnings at the New Business and Policy Refund stages.

• Distracted Driver Detection (Image Classification): Developed machine learning models to analyze drivers' activities and classify driving behaviors.

• Pure Premium Model (Insurance Pricing): Developed a generalized linear model for claim loss costs (pure premium) using Poisson and Gamma distributions for a P&C insurance company.

Cross Sell Model (Life Insurance):

• Cross Sell Propensity Model: Predicted the propensity of cross-sell opportunities using logistic regression.

• Next Best Product Model: Recommended next best products across categories for cross-selling.

• Agent Affinity Model & QSR Allocation Model: Evaluated customer-agent relationships and QSR allocations for orphaned customers.

Gramener (April 2017-November 2018);

• Led the Insights and Analytics Program: Delivered consumer insights, data sciences, and analytics across various domains, including Sales, Marketing, Digital, Product Planning, Corporate Strategy, and Innovation.

Selected AI/ML Projects:

• Key Driver Analysis (Customer Insights): Helped clients understand consumer purchase behavior and identify key improvement areas.

• Viewership/Impression Forecasting: Built causal/time series forecasting models using ARIMA, Holt-Winters, Random Forest, Neural Networks, and other machine learning algorithms.

• Brand Association Mapping (NLP): Developed an NLP engine to analyze brand perception, strengths, weaknesses, and competitor insights through social media data (e.g., tweets).

Previous Career Experiences

Cognizant Technology: Associate (July 2013 - May 2015)

Target Corporation: Associate Business Analyst (September 2011-June 2013)

Genpact: Business Analyst (December 2010 - August 2011)