Ashish P.

Technology Leader | Data Science | Artificial Intelligence | Machine Learning | Generative AI | Agentic AI | MLOps | Federated Learning | LVM | LLM | LMM | Tech Modernization | Enterprise Architecture | Cloud Deployment

Greater Philadelphia

About

I am subject matter expert and possess expertise in technology transformation, Artificial Intelligence, Data Science, Machine Learning, Cloud Migrations (from on-premises to cloud based Infra such as AWS, OCI etc and migrating large scale applications between cloud infra), application building and innovation. I have vast stretch of expertise and experience in domains like Finance, Insurance, Banking, Manufacturing, Construction with hands on experience building optimized solutions for Credit Card, Loans, Deposits, Securitization, POS, X-Sell, CRA, Partner Compensations including Launches, and product innovation. I am a subject matter expert of areas like Risk analytics and reporting. Excellent understanding on Regulatory reporting requirement wrt data and Risk catalogs (CCAR, DFAST, IFRS9, BCBS239, CCPA, CFPB and many more). Hands on involved in building new AI, ML & Data Science products using multiple algorithms and libraries for creating efficient solutioning for accurate forecasts with vision models and image detections for to support schedule and risk measurements. Good knowledge of tools like Oracle Primavera (P6), Aconex, Textura etc. I motivate my team members, creates a fun environments and care for them. I listen to their professional challenges and help find best ways to resolve them. Computer Vision: Image/Video Detection, Classification, Labeling & Tagging. Utilized CV2 and computer vision algorithms plus YOLO-v4-v11 with xx landmark points to recognize Expression and Emotion. Developed computer vision models for aerial, drone/UAV, satellite, EO/IR imagery. Developed Image processing engine using Large Image Models. LLM/RAG/Fine Tuning/Transformers: Experience Developing and implementing a comprehensive process for fine-tuning pre-trained large language models (LLMs) for specific generative tasks using the latest tools like PyTorch, Langchain, Langraph, Hugging Face Transformers, and the Llama library, ensuring a seamless adaptation to a targeted question-and-answer format. Built Interactive chatbots for multi-business & line support. Building Agentic Apps using MCP, A2A architecture embedded with Generative Ai and other Machine Learning Models including Classic ML as well as Deep Learning and other concepts including but not limited to Ontology and many more. Deep Learning: Experience in deep generative models including NLP, NLU, NLG, CNN, GANs, Diffusion, LSTM and Transformer models on images, tabular and text data. Contributions includes data augmentations, network design, loss function, training Strategies, and model evaluation.

Experience

  • Director - Data Science (Advanced Analytics, Research & AI Products Development) at Oracle
    Apr 2022 - Present · 4 yrs 3 mos

    As the Director & Head of Data Science for CIC Product(s), I lead the research, design, architecture, development, and modernization of next-generation software products across various business domains, including Construction and Utilities. My focus spans Computer Vision, Machine Learning, and Generative AI, leveraging cutting-edge technologies such as LLMs, LVMs, Whisper models, and YOLO (World and v8+) for image and video detection/classification. Building next-gen AI/ML applications for enhanced training, prediction, and forecasting, covering Classification, Regression, Clustering, Anomaly Detection, Labeling, Threshold Tuning. Additionally, NLP, OpenCV, CV2, CNNs, PyTorch, TensorFlow, Scikit-learn, Time Series Analysis including Explainability, Visualization, and scalable deployment. Creating Computer Vision models for Image/Video Detection, Classification, Labeling & Tagging. Developing prediction and classification tools using Logistic Regression, Random Forest, XGBoost, SVM, Bayesian inference, and other classical ML techniques Implementing Deep Learning models (NLP, NLU, NLG, CNNs, GANs, Diffusion) for data augmentation, network design, loss functions, training strategies, and model evaluation. Building and utilizing Large Language Models (LLMs) for text and speech processing in interactive applications. Implementing Fine-Tuning and Prompt Engineering to optimize AI-generated outcomes. Deploying and configuring LLMs using LangChain and other advanced frameworks. Leading the migration of AWS services to Oracle Cloud Infrastructure (OCI) for Data management, build automation, deployment, and AI/ML workloads along with Scalable, cost-effective cloud-based AI/ML solutions. Designing microservices architectures and leveraging Python, Scala, and cloud-native services. Build and deploy AI agents and Agentic AI apps using MCP/A2A architectures supporting hierarchical as well as agent mesh systems for critical applications and products.

  • Vice President - Data Services Analytics & Reporting at Barclays
    Sep 2019 - Apr 2022 · 2 yrs 8 mos

    Developed and tested new data representation methods for a machine learning model. Built and tested neural networks in PyTorch, including a multitask learning model incorporating LoRA. Created and lead multiple CD models under classification, regression, NLP, Deep Learning. Implemented algorithms to question users in order to accumulate data to improve training and inference. Senior Lead and Solutions Architect in Data Services, Financials, Risk & Regulatory Reporting. Technical expert for various use cases to solve business challenges in Risk & Finance for data services and reporting. Hands on technical experience working & Leading on Python, ML Packages, Analytics, notebooks and other data analytics libraries including Scripting, ETL & Big Data Technology like Hadoop, Hive, Impala, HBase, Mongo DB, Scala, Python, Spark, PySpark, Oracle PL/SQL, Shell Scripting. Good knowledge of AWS platform and services for data processing & transformations using Glue to Analytical services like Athena. Using S3 for data storage to Glacier for deep archives. Hands-on experience using reporting tools along with creating reports using Tableau, Qlik, MicroStrategy etc. Good understanding of Financials/GL & Credit Risk domain including but not limited to credit decisioning, Writeoffs, Impairments, Probability of Default(PD), Exposure At Default(EAD), Life Given Default(LGD), Capital and other models. Leading Data Services for Regulatory reporting & Compliance including but not limited to FDIC Reporting, FRY14M, CCPA, CCAR, DFAST, ALLL, IFRS9, CECL, AIRB, ForeST, & BCBS239. I am subject matter expert in these areas and posses expertise with legal and compliance aspects of this business area. I motivate my team members, creates a fun environments and care for them. I listen to their professional challenges and help find best ways to resolve them.

  • Barclaycard (Greater Philadelphia Area)
    • Assistant Vice President - Data Services Analytics & Reporting
      Dec 2016 - Sep 2019 · 2 yrs 10 mos

      Application development, Forecasting, Big Data Processing, Consumer Credit Risk Reporting, Capital model and Stress Testing. Hands on experience and expertise with regulatory model requirements, processing and reporting. CECL (Current Expected Credit Loss), CCAR (Comprehensive Capital Analysis and Review), DFAST, ALLL (Allowances for Loan Lean and Losses). Application development, Forecasting, Big Data Processing, Consumer Credit Risk Reporting, Capital model and Stress Testing. Hands on experience and expertise with regulatory model requirements, processing and reporting. CECL (Current Expected Credit Loss), CCAR (Comprehensive Capital Analysis and Review), DFAST, ALLL (Allowances for Loan Lean and Losses). Build pipelines Extract key compliance clauses and accounting terms from financial contracts, enhancing Barclays' ability to adhere to regulatory reporting standards (e.g., IFRS, FATCA). Implemented predictive models to flag high-risk transactions or clients from a compliance perspective (e.g., KYC/AML), reducing manual screening workload by 40% while maintaining regulatory accuracy. Collaborated with internal tax and finance departments to define machine learning features relevant to tax categories, cost allocations, and intercompany transfers, improving model interpretability for auditors. Utilized Barclays’ internal data lake to conduct time-series analysis on tax deferrals, earnings, and compliance trends across jurisdictions, using PySpark and Snowflake on cloud platforms. Contributed to the development of explainable AI (XAI) dashboards to support tax auditors and finance teams in understanding AI-driven risk scoring logic and audit recommendations.

    • Technical Lead - Data Services & Analytics
      Nov 2014 - Dec 2016 · 2 yrs 2 mos

      - Java/J2EE - Spark - Scala - Datastage - Python - PySpark - PL/SQL - Shell Scripting - MicroStrategy

  • Senior Technical Lead - Data Services & Analytics at Société Générale
    Apr 2014 - Nov 2014 · 8 mos

    - Java/J2EE - Oracle PL/SQL - MySQL - Tableau, Qlik - Shell Scripting - ab initio

  • Senior Developer Project Lead at TE Connectivity
    Oct 2013 - Apr 2014 · 7 mos

    - Java/J2EE - Oracle PL/SQL - Cognos - MySQL - Informatica - Shell Scripting - Jenkins CI/CD Pipeline