A

Antony Irudyaraj

About

Detail

Director of Data Analytics and AI (Principal Architect AI and Data Analytics) at Insurity
Texas, United States

Contact Antony regarding: 
work
Full-time jobs

Timeline


work
Job
folder
Project

Résumé


Jobs verified_user 0% verified
  • C
    Lead AI and Data Architect
    Connexus Credit Union
    Feb 2023 - Dec 2024 (1 year 11 months)
    Project Summary: Mortgage and Loan data Cloud Immigration Project Scope: I led the end-to-end transformation of Connexus Credit Union's mortgage and loan data infrastructure by architecting and executing a cloud-first, AI-powered modernization strategy. This initiative involved migrating legacy systems to a scalable AWS and Snowflake-based architecture, leveraging AWS S3 for data staging and AWS Glue for intelligent ETL pipelines. I unified disparate datasets
  • H
    Enterprise Data Architect
    HCL
    Apr 2020 - Feb 2023 (2 years 11 months)
    As the Director of Data, I oversaw the strategy, design, and execution of a comprehensive Customer Data Platform (CDP) aimed at delivering a 360-degree customer view and enabling targeted marketing initiatives. My responsibilities included directing the integration of data from multiple sources such as Salesforce, Experian, Google Analytics API, and SAP into an Azure-based architecture. I also led efforts to collaborate with the data science team in building advanced machine learning models for predictive insights and marketing automation. The project leveraged Azure Synapse and Databricks to build a robust data pipeline and ensure optimal data performance and security.
  • E
    Lead Data Architect
    Emids Technologies
    May 2019 - Apr 2020 (1 year)
  • Mphasis
    Data Architect
    Mphasis
    Aug 2018 - May 2019 (10 months)
    Key Responsibilities and Achievements: • Designed the Architecture and Data pipeline of this project. • Prepared Full Design Doc and Data model for this project. • Developed Sqoop scripts for initial loading of data • Performance tuning of Hive Queries • Improved Hive performance using Spark Hive context API • Developed Python scripts using Spark and Machine Learning libraries. • Ingest and Transform the data into Azure cloud platform • Compute the data using Snowflake cloud data ware house. • Wrote Shell scripts for data extracting from Adobe and injecting into HDFS. • Developed Spark transformations of Data • Actively participated in requirement gathering and analyzing business requirement. • Minimized the joins of Teradata in
  • Cognizant
    Technical / Data Architect
    Cognizant
    Feb 2012 - Aug 2018 (6 years 7 months)
  • Insurity
    Prinicipal Architect
    Insurity
    Key Responsibilities and Achievements: • LLM & Chatbot Development ◦ Directed the development of a self-service AI chatbot using AWS Bedrock Agents and Streamlit, allowing business users and underwriters to query structured and unstructured data using natural language prompts. ◦ Integrated Pinecone Vector DB and AWS Bedrock Knowledge Bases to enable retrieval augmented generation (RAG), ensuring LLM responses are grounded in enterprise specific documents. ◦ Addressed challenges in proprietary data licensing and vector storage compliance, ensuring secure and policy-aligned AI implementation. • External-Facing GenAI API Platform (FastAPI + Snowflake Cortex) ◦ Designed and deployed a multi-tenant GenAI platform using FastAPI to expos
Education verified_user 0% verified
  • T
    Diploma
    TRNP Polytechnic
  • Bharathiyar University
    Bachelor of Engineering
    Bharathiyar University
  • Microsoft
    Certified Artificial Intelligence (AI) Expert Microsoft Azure Certified : DP-900 - Azure Data Fundamentals AWS Certif
    Microsoft
  • I
    Implement
Projects (professional or personal) verified_user 0% verified
  • C
    CDP – for 360-Degree Customer View and Targeted Marketing
    Apr 2020 - Feb 2023 (2 years 11 months)
  • N
    NCAP Program
    May 2019 - Apr 2020 (1 year)
    The NexGen Cost Accounting Program (NCAP) within Kaiser Permanente will implement a costing systems and profitability reporting platforms with a single, consistent solution across Kaiser Permanente and its entities. The solution, a detailed-level cost accounting system which supports the integrated Kaiser Permanente health care model and encompasses the full continuum of care throughout all regions of Kaiser Permanente, will be utilized to support the allocation of membership and patient costing for health plans (in support of regulatory reporting and other operational measures), and utilization costing for owned hospitals and medical groups.
  • M
    MDA Program (Marketing and Digital Big Data Program)
    Aug 2018 - May 2019 (10 months)
    This Big Data analytics program to improve customer experience, marketing/sales efficiency, sales and retention across channels. This will require data from different data sources EDW (Tera data), MarketView (Oracle), Adobe analytics to ingest the data in Hortonworks Data lake to improve quality of data required for marketing analytics and build machine learning models. We are using Sqoop ingestion for extract the data from EDW (Tera data) and Oracle ingest into Hive table (RAW layer). We are using Unix scripts to Extract the data from Adobe Click stream and ingested to RAW Layer and build the model for (ETL Layer). In (Processing Layer) we are building the ML model and specific custom tables to apply the business logic for High growth cust
  • A
    AML SAM (Actimize Suspicious Activity Monitoring)
    Jun 2017 - Aug 2018 (1 year 3 months)
    The goal of this project is to rollout the strategic Actimize Suspicious Activity Monitoring (SAM V8) solution for the US Financial Services business. This will require the data sourcing of Client and Account Reference data as well details associated with the following types of transactions: Monetary Movements, Cash, Check, Wire, ACH, Bill Pay, Debit card purchases. Source files required for building these EDW table will be brought into Data refinery to As Is, Governed Gold layer and Provision Layer. As-IS Layer : Hadoop ingestion script will move files from Dropbox to HDFS Directories. Ingestion script will perform basic error and validation checks before copying data. Governed Layer : Data will be loaded in governed hive tables. HDFS dire
  • M
    MCR - (Main Frame Cost Reduction)
    Feb 2016 - May 2017 (1 year 4 months)
  • T
    Teradata Off-Load and Ab-Initio etl to Hadoop
    Oct 2015 - Feb 2016 (5 months)
    This project involved data being moved from Teradata legacy data-warehouse data using in-house artifacts and tools provided by Teradata itself. Tables in Teradata are of the size in excess of 10 Terabytes and each one of such tables has a huge amount
  • A
    Actavis-AZ Data Integration
    Apr 2015 - Oct 2015 (7 months)
    Astra Zeneca requires an extract of all COPD product sample transactions and their digital signatures in accordance with their required formats. Sample transactions for TD and DL exist across both Veeva and SAPCRM platforms since not all sample history exists in Veeva system. CASSRA transactions are included in scope as are the scanned copies of CASSRA forms which have signatures. Did POC of Data injection to HDFS.
  • E
    ECC Phase 3 and Phase 4
    Nov 2014 - Apr 2015 (6 months)
    Electronic Customer Communications program is the FPL's strategic initiatives to build the enhanced system whose ultimate goal is to create a solid framework for ECC and implement and conduct a live end-to-end test of the redesigned customer experience for customers enrolled on EMB, as well as those choosing to view and/or pay their bill on FPL.com. Focus on building the framework to support core ECC functionality
  • O
    Out Bound Process
    This MCR project used for the cost reduction of the Main frame system and other source systems in Navient. We used Syncsort DMX-h tool for offloading the data from Mainframe to Hive in Hadoop platform. This project has two process. 1. This process offloading data's from DB2 to Hive staging tables in Hadoop platform using Cobol copy book and Syncsort-DMX-h tool. Hive staging to Hive tables we used Unix bash scripting for inject the data to Hadoop. 2. This process migrating the existing SAS programs and Mainframe JCL jobs to Hadoop platform using Syncsort – DMX-h tool. It's internally using Hive queries and SAS business logic for this migration process. For SQLServer Offloading, we used Syncsort DMX-h tool for inject the data to Hive and used
  • a
    and AI
  • E
    Enterprise-level
  • C
    Cloud Migration
This is a community-created genome.