Generative AI Engineer and MLOps Specialist

VERONICA ALURI

AI Engineer *****************@*****.*** +1-312-***-**** LinkedIn : linkedin.com/in/veronica-aluri-385490414

Professional Summary:

Results-driven Generative AI and Machine Learning professional with over 5 years of progressive experience building intelligent, data-driven solutions across enterprise environments, spanning the full spectrum from data analytics and predictive modeling to production-grade LLM and RAG application development.

Hands-on expertise developing and deploying Generative AI applications using Azure OpenAI, GPT-4, and GPT-4o, with deep practical experience in RAG pipelines, LangChain agent workflows, Prompt Engineering, and vector database integration using Pinecone and Weaviate.

Proven ability to build end-to-end machine learning pipelines from raw data ingestion through model training, evaluation, and production deployment, leveraging PyTorch, TensorFlow, XGBoost, Scikit-learn, boto3 and HuggingFace Transformers across complex enterprise datasets.

Strong background in NLP and Transformer-based model development, with hands-on experience fine-tuning BERT, MiniLM, T5, and SBERT for real-world tasks including text classification, question answering, named entity recognition, and intelligent document processing.

Experienced in designing and delivering scalable MLOps infrastructure using Docker, Kubernetes, MLflow, Azure ML, and CI/CD pipelines, ensuring AI systems are production-hardened, version-controlled, and continuously monitored for performance and reliability.

Proficient in processing and transforming large-scale datasets using PySpark, Databricks, Snowflake, SQL, and Apache Airflow, building robust data foundations that feed high-quality inputs into downstream AI and machine learning workflows.

Solid foundation in data analytics and business intelligence, with demonstrated experience building advanced Power BI dashboards, engineering DAX models, authoring complex SQL queries, and translating raw enterprise data into actionable insights for business stakeholders.

Collaborative and technically versatile professional with a strong track record of working across cross-functional engineering, cloud, and business teams to deliver AI-powered solutions that reduce manual effort, improve operational efficiency, and drive meaningful business outcomes through emerging technologies.

Technical Skills:

Generative AI & LLMs Azure OpenAI, GPT-3.5, GPT-4, GPT-4o, LLaMA, LangChain, Agentic AI, LLM Orchestration, RAG, Prompt Engineering, Workflow Automation, Few-Shot Learning, Chain-of-Thought Reasoning

NLP & Transformer Models BERT, MiniLM, T5, SBERT, HuggingFace Transformers, spaCy, Text Classification, Named Entity Recognition, Document Intelligence, Sentence Segmentation, Entity Extraction

Machine Learning & Deep Learning PyTorch, TensorFlow, Keras, Scikit-learn, XGBoost, Random Forest, OpenCV, SHAP, LIME, A/B Testing

Vector Databases & Semantic Search Pinecone, Weaviate, FAISS, Vector Embeddings, Semantic Search

Data Engineering & Big Data PySpark, Databricks, Delta Lake, Apache Airflow, Azure Data Factory, Snowflake, Hive, Pandas, NumPy, SQL

MLOps, Cloud & Infrastructure MLflow, Docker, Kubernetes, Azure ML, AWS SageMaker, GCP Vertex AI, Terraform, GitHub Actions, Azure DevOps, Jenkins, AI Observability, Logging, Metrics, Tracing

Analytics, Backend & Visualization Python, R, FastAPI, Power BI, DAX, Power Query, Excel, Power Automate.

Responsible AI & Governance Responsible AI, AI Guardrails, ML Governance,Secure AI Systems, AI Compliance, Risk Management

Educational Details:

Governors State University - Illinois, USA

Master of Science in Computer Science Aug 24 - Apr 26

Annamacharya Institute of Technology and Sciences (AITS)

Bachelor of Technology in Civil Engineering May 15 - Mar 19

Professional Experience

Capital One Chicago, IL

Gen AI Developer Mar 2025 – Present

Engineered production-ready Generative AI solutions powered by Azure OpenAI, GPT-3.5, GPT-4, and GPT-4o that transformed how enterprise teams interact with massive volumes of business data, enabling intelligent document understanding, on-demand summarization, and precise context-aware query resolution at scale.

Constructed sophisticated RAG (Retrieval-Augmented Generation) pipelines that bridge the gap between static enterprise knowledge bases and dynamic Large Language Models, dramatically improving response accuracy, contextual depth, and factual grounding in AI-driven business applications.

Developed high-performance semantic search capabilities by architecting embedding workflows with Pinecone and Weaviate vector databases, enabling users to surface the most relevant information from vast document repositories through natural language queries with exceptional speed and precision.

Built intelligent, multi-step AI agents and reasoning workflows using LangChain, incorporating prompt orchestration, persistent memory, and dynamic tool chaining to handle complex, multi-turn interactions that go far beyond simple question-and-answer exchanges in enterprise environments.

Crafted and continuously refined Prompt Engineering strategies including few-shot learning, chain-of-thought reasoning, and instruction tuning, directly elevating the reliability, coherence, and business alignment of responses generated by deployed LLMs across varied enterprise scenarios.

Fine-tuned domain-adapted Transformer-based NLP models including BERT and MiniLM through the HuggingFace Transformers library, pushing model performance on critical tasks such as question answering, named entity recognition, text classification, and automated document intelligence well beyond out-of-the-box baselines.

Designed and delivered robust backend AI services using Python and FastAPI, translating complex Generative AI capabilities including summarization, semantic retrieval, and conversational query resolution into clean, secure, and low-latency APIs seamlessly consumed by enterprise-facing applications.

Built end-to-end data ingestion and preprocessing pipelines using PySpark, applying targeted NLP techniques including text normalization, sentence segmentation, and entity extraction to ensure high-quality, well-structured inputs that maximize the reasoning accuracy of downstream LLM and Transformer models.

Leveraged PyTorch and Scikit-learn to drive rigorous model experimentation, feature engineering, and performance benchmarking, validating and strengthening the predictive components that complement core Generative AI capabilities within broader enterprise machine learning ecosystems.

Packaged AI inference services into portable, production-hardened containers using Docker and orchestrated their deployment at scale through Kubernetes, ensuring consistent runtime behavior, fault tolerance, and seamless scaling of model workloads across all environments.

Embedded MLflow deeply into the machine learning development workflow to bring full transparency and control over experiment tracking, model versioning, and performance benchmarking, enabling the team to iterate rapidly and ship reliable Generative AI pipelines with confidence.

Designed and supported real-time event-driven data pipelines using Kafka, RabbitMQ, Azure Event Hub, and Pub/Sub messaging systems to enable scalable AI and data processing workflows.

Partnered with cloud and engineering teams on Azure infrastructure to ensure AI applications are deployed securely and scale gracefully under real-world demand, while proactively instrumenting observability layers to monitor inference latency, catch input anomalies early, and drive continuous model reliability improvements.

•. Implemented AI guardrails, observability, and monitoring frameworks using logging, metrics, and tracing to improve reliability, security, and compliance of production Generative AI systems.

Capgemini Bangalore, IN

Project Client: Property & Casualty (P&C) Insurance

ML Engineer / AI Engineer May 2022 – July 2024

Designed and delivered production-grade machine learning solutions for predictive modeling, classification, clustering, and recommendation tasks, developing and validating models using XGBoost, Random Forest, TensorFlow, boto3 and PyTorch with end-to-end implementation in Python and R.

Built end-to-end data science pipelines covering data extraction, feature engineering, model training, and evaluation, writing scalable transformation logic in PySpark and SQL, orchestrating workflows through Databricks, and tracking experiments using MLflow.

Developed and fine-tuned NLP and transformer-based models for text classification, summarization, question answering, and behavior prediction, leveraging BERT, T5, and SBERT through Hugging Face Transformers and applying linguistic preprocessing with spaCy.

Built and operated batch and real-time inference pipelines to support high-volume model serving and low-latency feature computation, engineering data movement through Azure Data Factory, processing distributed workloads using PySpark, and managing model lifecycle with MLflow.

Processed and managed large-scale structured and unstructured data to support downstream model training and analytics, performing high-performance querying through Snowflake and Hive, building ETL pipelines using Apache Airflow, and handling data transformation with Pandas and NumPy.

Established MLOps practices by containerizing model environments using Docker, managing deployment orchestration through Kubernetes, and implementing automated CI/CD workflows for versioned and reproducible model releases using Azure DevOps, Jenkins, and GitHub Actions.

Optimized ML training and inference workloads on cloud platforms by running scalable model jobs on AWS SageMaker and GCP Vertex AI, managing compute infrastructure through Azure ML and EC2, and automating resource provisioning using Lambda, Cloud Functions, and Terraform.

Built and maintained streaming and message-based data pipelines using Kafka, RabbitMQ, Azure Event Hub, and Pub/Sub architectures to support distributed ML workflows and real-time data processing.

Developed model validation and explainability frameworks to interpret and audit model predictions using SHAP and LIME, and implemented A/B testing and statistical experimentation frameworks to measure model impact and validate AI-driven strategies across production environments.

Designed and trained deep learning and computer vision models for image classification, object detection, and document parsing using TensorFlow, PyTorch, Keras, and OpenCV, while implementing incremental CDC batch pipelines and Change Data Feed sync patterns through Databricks Delta Lake for real-time feature computation.

Implemented incremental data sync and real-time feature computation patterns using Databricks Delta Lake and Change Data Feed, supporting low-latency ML model inference and keeping feature stores consistently updated across training and serving environments.

Beginning in 2023, explored and evaluated early-stage large language models including OpenAI GPT, Azure OpenAI, and LLaMA for potential workflow automation use cases, experimenting with zero-shot and few-shot prompting techniques to assess output quality and domain alignment for internal proof-of-concept initiatives.

Beginning in 2023, explored Retrieval-Augmented Generation concepts by conducting early proof-of-concept work integrating language models with vector stores such as FAISS and Pinecone using LangChain, evaluating retrieval accuracy and response quality for potential future production use cases.

Designed and trained deep learning models for document parsing and unstructured data extraction, processing raw inputs through PyTorch and TensorFlow, and built automated ETL pipelines using Apache Airflow and Python to feed cleaned and structured data into downstream model training workflows.

Wipro Bangalore, IN

Project Client: Target

ML Engineer/ Data Scientist Nov 2019 – Apr 2022

Designed and developed scalable machine learning solutions using Python, leveraging Scikit-learn, XGBoost, BERT, T5, and SBERT to build intelligent document classification, semantic search, and enterprise knowledge management applications.

Developed end-to-end machine learning and NLP applications using Python, implementing feature engineering, model optimization, and REST API integration with Flask and FastAPI for production-ready services.

Fine-tuned transformer-based language models using PyTorch, TensorFlow, Hugging Face Transformers, spaCy, and NLTK to improve text classification, document summarization, semantic search, question answering, and information extraction capabilities.

Designed and implemented intelligent document search and semantic similarity pipelines using Word2Vec, TF-IDF, and FAISS, enabling context-aware document retrieval and accurate matching across enterprise datasets.

Built scalable machine learning pipelines using GCP AI Platform, Google Cloud Dataflow, and Scikit-learn, automating data ingestion, feature engineering, model training, hyperparameter tuning, and predictive analytics for large-scale datasets.

Developed and deployed machine learning solutions on Amazon SageMaker, EC2, AWS Lambda, Amazon S3, AWS Glue, and Amazon Athena, implementing scalable inference pipelines, automated workflows, cloud monitoring with CloudWatch, CI/CD using CodePipeline and CodeBuild, and infrastructure provisioning through Terraform.

Engineered enterprise data pipelines using Apache Airflow, Informatica, MS SQL Server, and Snowflake, creating automated ETL workflows and optimized data warehouses to support machine learning model training, business intelligence, and advanced analytics.

Developed deep learning models using GPU acceleration and CUDA, implementing Autoencoders and Variational Autoencoders (VAE) for anomaly detection, dimensionality reduction, feature learning, and high-performance model inference.

Designed interactive web applications and model monitoring dashboards using React.js and JavaScript, providing real-time visualization of model performance, prediction analytics, operational metrics, and system health.

Containerized machine learning applications using Docker and orchestrated scalable deployments with Kubernetes, implementing automated CI/CD pipelines using Jenkins and Azure DevOps to ensure high availability, fault tolerance, and reliable production deployments.

Built scalable ML platform integrations using Apache Airflow, Python, and SQL, enabling automated feature pipelines, batch model scoring, and data transformation, while developing rule-based chatbots and interactive analytics dashboards using Rasa, Dialogflow, and Streamlit.

Contact this candidate