• Built a bio medical products Omni channel buying recommendation engine using collaborative filtering and deep learning-DNN/RNN/CNN based models on google cloud platform to recommend 10k products on the e commerce website• Built a Lifesciences browsing focused on Bayesian personalized ranking recommendation engine model Bigquery ML leading to a personalized recommendations for the users.• Built a customer propensity machine learning model to predict the likelihood of user buying a product online for lifts in user engagements. • Built a browsing purchase path analysis and customer segmentation modelling using markov chain and Bisect K means. • Built a custom generative AI/LLM chatbot to improve product search for 4000 products streamlining customer support process. • Created a unified customer profile/ digital twin of sales users using ML solutions, including products affinity, purchase similarity, CLTV, RFM generating 100% visibility of users for targeted marketing include customer clustering based on purchases, leading to 100% customer account coverage.• Architected a Microsoft graph API based data service to retrieve, transform and show customer insight on Marketing engagements • Designed ETLs in AWS and GCP environments to ingest 200 GB data and integrated multiple APIs for google CM360, Microsoft graph API, GA4 API, Atlassian cloud and on-prem APIs and saved $1.1M in infra and labor cost. • Mentored data engineers, data scientists and analysts to setup Data warehouse and ETL framework supporting multi cloud environment at AWS and GCP.GEN AI: OLLAMA -CODESTRAL, Annoy and Streamlit powered RAG, Built On Prem Chatbots with API EndpointsAPI Exposure: Graph API, GA4 Reporting API, Google Ads API, Google Campaign manager 360 API, AWS S3 API, Jira API,DialogFlow Cx API, Flask Framework, ON24 Event API

• Designing algorithms for warehouse optimization using mixed-integer linear equation/linear equations.• Designing statistical models for inventory levels and cost predictions using JMP/R/Python.• Built a Warehouse part shortage prediction machine learning tool resulting in a daily prediction/binary classification with 70% or more precision on a large unordered data set using save the best result approach from random forest/SVM/Adaboost and logistical regression algorithms.• Text prediction using RNN/LSTM neural network/deep learning for warehouse job precedence using tensor flow, Keras, pickle libraries and using select the best approach with low loss and high accuracy.• Text extraction from a highly unordered data set and periodic report automation leading to 100% reduction in manual work using python libraries like xlsxwriter, NumPy, regex, pandas.• Deployment of machine learning models on Azure Cloud using DataIKU/Databricks and containerization using Kubernetes/docker and Mleaps, heroku and flask. • Built metrics to increase visibility of ship side action tracker (SAT) to measure Kit Quality, Material integration Center delivery, SAT Reason Codes trackers, improving to 100% automation and visibilty of SATs.• Built Part Re-shuffling and inventory recommendation engine in Python for 3D Robotic warehouse with Shortest path selection.Project Administration/ML ops• Administrating Team foundation server and git repositories for periodic changes.• Deploying R shiny Apps for warehouse inventory signals using SQL, R shiny, CSS code.• Identifying areas in warehouses where AI/ machine learning can be used in the supply chain.• Building ETL and ML Pipelines using cloud/compute functions, integrating BI engine with ML model output

Led a high stake, high impact, high visibility warehouse optimization and automation algorithm design task using linear, mixed integer linear programming,excel solver, python linear solvers. Conducted Data Mining and ETL using python,SQL/SSIS for Warehouse Optimization on large dataset.Performed a Statistical analysis using Python, R, and Tableau of the warehouse business problem.Created a Warehouse KPI Dashboard for measuring various warehouse elements, using SQL and Tableau.Applied Clustering algorithm to find the optimal location for parts in a 3 dimensional warehouse setup.

1. Recipient of dean's merit scholarship machine learning: ML models:K-Nearest Neighbor, Linear and Logistic Regression, Neural Networks, Lasso, Feature Selection, Support Vector, Machines, Decision Trees, Boosting, Naive Bayes, Gaussian Processes, k-Means Clustering, PCA, NMF,Word2Vec, MDP, DDPGML Paradigms:Instance-Based Rule-BasedOptimization Probabilistic

• SPOC for Blackberry Z10 and Q10 Apps Development, deployment, publishing using QT, C++,Microsoft SQL Server, Performance Tuning, Query Profiling and optimization.• Lead Reporting and analytics tool Development in C++, SQL for Memory profiling, Performance improvement, User experience using Embarcadero RAD Studio XE4/XE3 and SQL Server 2008.• Enabled Application for Unicode capable multi-character/multilingual input and searches.• Built New customized library component(s) to be used across an application for RAD Studio XE4.• Added new functionality in the embarcadero RAD Studio XE 4 IDE, for Hotfixes provided by Embarcadero and redeployment for Developers.• Performed Data Analytics and Reporting using Microsoft reporting services and Auto Tag using MS SQL reporting engine, Winward reporting.• Lead the reporting and analytics module having more than 2700 reports, their upgrade on time and with accountability.

• Technical implementation and management of ATM/POS/EFT Switch to gateway interface for Canadian and US banking systems.• Interface and UI development of the card management system (CMS) development as part of a large retail banking suite.• Core expertise in Credit/Debit/Value cards and ISO 8583, ISO-7812, ISO -3166• Worked on a wide variety of ISO 8583 based messaging for debit cards.• Built Knowledge over various retail banking card products and banking system, payment gateways, payment switches architecture both front end, back end.• Developed expertise over entire life cycle of bank cards i.e. ordering,creation,authorization,activation,dispatch,alteration,replacement,bulk ordering,Deactivation,Re-Activation.• Configured Payment gateway registration, switch integration, ATM, POS.MOTO, IVR integration.• -Workflow designing/Development for Card modules, gateways and simulator development.• Involved in a major bulk magnetic card replacement with Chip Cards.• Developed XML/XSLT based interface/Parsers for Lending applications used by credit rating agencies.

• Worked as a Designer, Developer, and mentor for 50 successful minor and major releases of the In House Product Development. • Involved in requirement discussion, Design, Documentation, Front end Coding, Testing, Build, Packaging, Back-end coding of the product.• Been an Effective coder in C++/UI Developer and in optimized SQL.• Expertise in SVN, JIRA (Admin Level), Confluence (Admin Level)• Data Mapping tool development and usage from old to new system for large data migration.• Doxygen, UML, Build, Code Counters, Optimization of SQL and Writing Complex Searches.• Developed Identity access matrix (IAM) generation on SQL Server across 3 data centers.• Data Concurrency identification and resolution using SQL triggers.• Pattern Algorithms and their usage on a large array of data.

• From the Scratch Design and Analysis of requirements of Interfaces for System Integration.• Taken initiative to integrate payment gateway with payment switch.• Worked closely with Architects, Director Technology and other product owners From ATLANTA, GA for requirement gathering and system integration and project coordination.• Created Framework for ease of development, Perl scripts for application Analysis.• Data Packet Size analysis to reduce network load, saving every single space of byte.• Responsible for design, development, implementation of ISO 8583, protocol messages for first data debit gateway network.• Implemented packet data compression algorithms quiet effectively.• Successful implementation/Integration of ISO 8583, 7812 based interface for ATM/IVR/MOTO/POS for STAR, PULSE, NYCE network in an absolutely isolated and zero assistance scenario.• ISO 3166 integration with all three networks.• Independent switch simulator development for testing purposes by creating message parsers for packed and unpacked EBCDIC, ASCII, HEX data content.• Hex Dump Analysis and debugging for generating a high throughput transaction efficiency which also included a time space trade off analysis to attain max transactions per second (TPS) with a least usage of memory.• Handled a very Organized and Neat Design documentation and Large re-engineering initiative.• Developed a Very stable ISO 8583 bidirectional message parser for financial transaction and successfully integrated the new system with various credit and debit card networks.• Language/ENV used: C, UNIX. Stratus VOS.

--Airbus,GMBH-Developed a BITE(Built in test equipment ) for A380 Passenger aircraft , one of the Biggest and most complex module of aircraft ,handling interaction with multiple in flight director board, and other devices mounted on aircraft like in flight entertainment,illumination system,landing gear system,door opening system etc. Skils used:C on SUSE Linux /windows with config management on WINCVS, Clear case and Clear Quest, DOORS, RT tester and Trace32,Artisan ,this involves debugging ,writing Requirement Specs,SRD,SDD,SCS,Test cases etc.--- Nortel Networks,Canada :Lawful Intercept(Product Development and support )• Handled Nortel Lawful Intercept Modules in a multilayered complex architecture for Telephone call interception and CS2k SoftSwitch for VOIP telephony.• Owned Redirect media gateway controller module for gateway interchange messages.• Involved in major multilayered upgrades in BCT (bearer channel tandeming).• Effective coder and proactive team member of Nortel Lawful intercept Team.• Depth Knowledge of SDP,SIP,MGCP,IPSec,ISDN,ISUP,H.264,Call tracing,3WC and telecom RFC's provided from IETF(internet engineering task force).• Real Life setting of Telephone interception debugging at china, russia, Malaysia, turkey, israel and Canadian switches .• Tools used : C++/win32,COM,ASP,MS-access, windows web server 2003,WINCVS,OOAD UML with VUML 3.0,.NET SDK 2003,HP-UNIX and Clarify clear support for defect control system.