Srinivasa Gadireddy
AI Architect
Mckinney,TX
Summary
Dynamic AI Architect and AWS Certified Machine Learning Specialist with over 20 years of IT experience, including 8 years focused on Machine Learning. Proven track record of delivering innovative solutions that address complex business challenges through advanced data mining, statistical analysis, and machine learning techniques. Expertise in Deep Neural Networks (MLPs, CNNs, RNNs, Transformers) using TensorFlow and Keras, achieving a 15% improvement in investment returns for clients. Skilled in developing scalable models and frameworks that integrate seamlessly into existing products while fostering collaboration across organizations to enhance product design and processes through data science applications.
Overview
21
21
years of professional experience
5
5
Certifications
Work History
Cloud Architect / Lead Software Engineer
Guardian Life Insurance
NJ
04.2022 - Current
- Platform-focused Cloud Architect leading large-scale identity, security, and application modernization initiatives supporting Guardian Anytime customer experiences, with a strong emphasis on scalability, security, and future AI/ML enablement.
- Owned and led the Guardian Anytime Registration and Login platform, supporting 1.5M+ new customer registrations with high availability, security, and performance at scale.
- Led the enterprise security migration from LDAP to Okta, redesigning authentication and authorization flows to align with modern Zero Trust and cloud-native identity patterns.
- Partnered closely with Cybersecurity and IAM teams to embed security controls directly into platform architecture and application design.
- Modernized identity and access layers to ensure AI/ML and GenAI readiness, enabling secure model access, user context propagation, and future agent-based workflows.
- Designed and governed cloud-native architectures leveraging Kubernetes, PaaS services, APIs, and managed databases, following enterprise best practices for scalability and resilience.
- Defined and authored REST API contracts, integration standards, and operational guidelines for configuration, maintenance, and long-term platform support.
- Built and reviewed frontend application interfaces using React.js, ensuring seamless integration with backend identity and platform services.
- Delivered application and service proofs of concept, supported pilot launches, and iteratively improved solutions based on stakeholder and user feedback.
- Played a key role in platform upgrades and stability initiatives, ensuring minimal downtime and secure evolution of critical customer-facing systems.
- Recognized as a top performer for consistent execution, ownership, and cross-team leadership.
- Environment: Spring boot, OKTA, OAG, OpenSearch, S3, ECS, Reach/Nextjs, Kubernetes/EKS
AI/ML Specialist Solution Architect
Axa Equitable
Newport, NJ
09.2017 - 03.2022
- Pegasus Agent Studio – Enterprise GenAI Agentic Workflow Platform: Led the architecture and deployment of Pegasus Agent Studio, enabling enterprise teams to build and execute GenAI-powered agentic workflows using frameworks like LangChain, LangFuse, and Flowise. Enhanced platform capabilities to support Retriever-based RAG agents, multi-model routing, and full observability for every agent run. Enabled MCP server integration for bidirectional workflow triggers, connecting AI agents with enterprise automation pipelines.
- Auto Insurance Payout Estimator: Part of automating the auto insurance claim process, Deep learning service is developed based on MASK R-CNN to estimate vehicle high level damages and to predict near realtime payout amount.
- Topic modeling and Chatbot: Developed NLP based tools to automatically tag customer support tickets according to topic, or recognizing patterns and routing conversations to the most appropriate team. Responsible for Intent based chatbot using RASA framework ( used Keras Policy based on LSTM). Later updated to a fine tuned BERT model.
- Insurance FraudDetection: Fraud detection is a cognitive application to predict fraud events, detect fraudulent activity, suspicious claims, and subtle behavior patterns using multiple techniques.
- Responsibilities:
- Developed bidirectional workflow triggers enabling Pegasus workflows to call agentic flows and vice versa.
- Responsible for data aggregation, data pre-processing, missing value imputation and descriptive and inferential analysis.
- Tackled highly imbalanced dataset using sampling techniques like down-sampling, up-sampling and SMOTE (Synthetic Minority Over-Sampling Technique)
- Developed event-driven architecture and presented to leadership. S3, Lambda, Kinesis Firehose, sagemaker. Custom ETL jobs in Glue using Python, PySpark and boto3.
- Supporting production implementation activities to predict the probability of fraud in real time.
- Contribution to designing and implementing a web application development frameworks that allow to efficiently develop and deploy web applications that can also serve machine learning algorithms.
- Experience in hypothesis testing, multivariate testing (A/B testing) and optimization algorithms, descriptive, exploratory, inferential, predictive modeling of the given scenario along with data modeling, and recognizing key performance indicators (KPI)
- Deliver and drive model/analytics insights throughout organization, helping to support data-driven decision making at all levels of the organization
- Experience with pre deployment optimization like model quantization and post model deployment support like concept and data drift.
- Experience with kubeflow pipelines, kubeflow and sagemaker integrated pipelines
- Environment: Python, LangChain, , Hadoop, Redshift, S3, Kinesis, Mongo 3.x, GIT, AWS Sagemaker, AWS boto3, Kafka, Hadoop, Hive, random forest, Kubeflow, HDFS, SVM, MapReduce, Tensorflow 1.x
Sr Big Data Developer/ Data Scientist
Axa Equitable
Newport, NJ
09.2014 - 08.2017
- Customer 360 is a customer dashboard aggregating all customer insights, personalize marketing, customer segmentation and lifetime value prediction for sales persons/advisors.
- Responsibilities:
- Involved in creating Data Lake by extracting customer's Big Data from various data sources into Hadoop HDFS. This included data from Excel, Flat Files, Oracle, SQL Server, MongoDb, HBase, Netezza and also log data from servers.
- Designed and provisioned the platform architecture to execute Hadoop and machine learning use cases under Cloud infrastructure, AWS, EMR, and S3.
- Worked with parameter tuning and model evaluation techniques Confusion Matrix, Cross validation, AUC-ROC etc. Customer Profiling models using K-means and K-means++ clustering algorithms to enable targeted marketing.
- For the customer life time value problem, calculated the errors using various machine learning algorithms such as Linear Regression, KNN, DecisionTreeRegressor, SVM, Bagging Decision Trees, Random Forest, AdaBoost, and XGBoost. Chose the best model eventually based on MAE.
- Expertise in leveraging the Exploratory Data Analysis (EDA) with all numerical computations and by plotting all kinds of relevant visualizations to do feature engineering and to get feature importance
- Environment: Python, SQL, Oracle 12c, Netezza, SQL Server, Redshift, Java, PL/SQL, Tableau, MLlib, Kafka, Hadoop Ecosystem 1.X / 2.X, MapReduce, Hive 0.11, HBase 0.9, AWS DevOps, SVM, Tableau, XML, AWS EMR, GIT.
Application Architect
Axa Equitable
Newport, NJ
07.2011 - 08.2014
- ONE is Intranet project to rewrite the current AXA Equitable intranet components to align with the new AXA group intranet standards and provide AXA a new Global Intranet platform.
- Responsibilities:
- Implemented Remote portlet consumption using Iframe
- Experience creating Quickr place template, customizing the REST responseGood experience Installing, fixpacks on IBM websphere portal server..
- Ajax functionality is developed using JQuery.
- Environment: J2EE6,JSF2.0, Hibernate 3.0, Jquery1.8, RAD7.5, REST, WAS 7.0, W.P.S 7.0,W.C.M 7.0 ,Quickr 8.5,Connection 2.5,Oracle 11g, ANT 1.5, Log4J, JavaScript, XML, Selenium.
Sr. Java server faces developer
Princeton University
Princeton, NJ
12.2010 - 06.2011
- LA is is a web based application used by the Princeton University internal users . The LA system is centrally located between the University's other large institutional ERP systems interfacing and integrating PeopleSoft HR, General Ledger and Budget,and Loans and Receivables on the other side.
- Environment: J2EE6,JSF2.0, EJB3.1, Hibernate 3.0, Eclipse3.6,Glassfish3.0,JIRA 4.1,Oracle 11g, ANT 1.5, Log4J, AJAX,Jquery1.2, JMS, JavaScript, XML,MS Visio, Selenium, Web Services, SVN, CSS.
Sr. Java Web Lead
BlueShield of California
San Francisco, CA
10.2009 - 11.2010
- Employer Self Service (ESS) is a web based application of BlueShield to it’s clients. It gives Ability for Group Administrators and Producers to view member level eligibility and high level benefit information online. Project scope extends to all products and all LOBs. ESS application enables employer users, as well as producers and internal BSC users, to manage user activations and access rights.
- Environment: JSF, Spring, AJAX, DOJO, JSP, JavaScript, Oracle, , RAD 7.0 (Web Sphere 7.0), Web Sphere Portal server 6.1 WCM6.1, Web Services, Log4J, PVCS, JUNIT,CSS.
Sr. Java Web Interface Developer
Infocrossing Healthcare services
Jefferson City, MO
07.2008 - 09.2009
- Emomed is a web-based Missouri state Claim Management Medicaid portal which encompasses Web Portal, various Web Applications, and Enterprise Service Bus. Medicaid enrolled providers can use this portal for submitting claims, registration of users .Providers can also get the latest news feeds from Missouri Medicaid.
- Environment: JSF, AJAX, DWR, JSP, JavaScript, DB2, Spring 2.0, RAD 7.5 (Web Sphere 7.0), Web Sphere Portal server 6.1, WCM 5.1, Web Services, CICS 3.1,WSRP, Log4J, SVN, JUNIT,CSS.
Sr. Java Web Interface Developer
Tata Consultancy Services (TCS)
Mumbai
03.2007 - 06.2008
- Developed international fares engine. Fares are calculated based on set of rules as a nightly job. All fares are uploaded to Airline Tariff Publishing Company (ATPCO). ATPCO provides fare data in an electronic format. End users for the fares are online travel agencies such as Expedia, Orbitz, and Travelocity.
- Environment: Struts 1.2, AJAX, DOJO, JSP, JavaScript, Oracle 10g, SQL/PL-SQL, Spring 2.0, RAD 7.0 (Web Sphere 6.1), Web Services,WCM5.1, WSRP, Log4J, Clear Case,CSS, JUNIT, iReports.
Java Developer
Tata Consultancy Services (TCS)
Mumbai
11.2004 - 02.2007
- Supply Chain Management System, the project involves design and development of an Internet based Online Supplier chain management System to streamline the operations of Suppliers and Company. The aim of the system is to tightly integrate the various operations of the suppliers and company into a web enabled transaction environment.
- Environment: JSP, JavaScript, Oracle 9i, SQL, XML, XSLT.
Education
B.E - Electronics and Communication
Jayam College of Engg and Tech
India01.2004
Skills
- Proficient in Python, Java, and JavaScript
- Cloud management expertise
- Experience with Keras and TensorFlow in deep learning projects
- Machine learning techniques
Certification
Tensorflow Certified Developer
Editors
Sagemaker studio, Jupiter, PyCharm , VSCode and Eclipse.
Timeline
Cloud Architect / Lead Software Engineer
Guardian Life Insurance
04.2022 - Current
AI/ML Specialist Solution Architect
Axa Equitable
09.2017 - 03.2022
Sr Big Data Developer/ Data Scientist
Axa Equitable
09.2014 - 08.2017
Application Architect
Axa Equitable
07.2011 - 08.2014
Sr. Java server faces developer
Princeton University
12.2010 - 06.2011
Sr. Java Web Lead
BlueShield of California
10.2009 - 11.2010
Sr. Java Web Interface Developer
Infocrossing Healthcare services
07.2008 - 09.2009
Sr. Java Web Interface Developer
Tata Consultancy Services (TCS)
03.2007 - 06.2008
Java Developer
Tata Consultancy Services (TCS)
11.2004 - 02.2007
B.E - Electronics and Communication
Jayam College of Engg and Tech