Home

Achyutha Pranavi - AI/ML ENGINEER

[email protected]

Location: Jersey City, New Jersey, USA

Relocation: REMOTE

Visa: GC

Resume file: Achyutha_Pranavi_Senior_AI_ML_Engineer_Resume_1754412612856.pdf Please check the file(s) for viruses. Files are checked manually and then made available for download.

Name: ACHYUTHA PRANAVI Role: Senior AI/ML Engineer | Data Analyst Email address: [email protected] Contact: +13479210136 LinkedIn: http://www.linkedin.com/in/achyutha-p-89a3b5216 Professional Summary Results-driven Senior AI/ML Engineer with 11 years of progressive experience in data science, machine learning, and AI solutions across finance, healthcare, retail, insurance, and telecom sectors, delivering measurable business impact through advanced analytics and AI implementations. Expert in cloud-native AI/ML solutions on both Microsoft Azure (Azure ML, Databricks, Synapse Analytics, Cognitive Search, Azure OpenAI) and AWS (Redshift, S3, Lambda, Athena, Glue, QuickSight), with proven ability to architect and deploy scalable, production-ready systems. Advanced practitioner in Large Language Models, Cursor AI and Generative AI, building retrieval-augmented generation (RAG) pipelines using LangChain, Semantic Kernel, FAISS, and Azure Cognitive Search to reduce hallucinations and enhance accuracy in knowledge-heavy financial and healthcare contexts. Developed transformer-based NLP models using BERT, GPT, and Azure OpenAI for financial document summarization, sentiment analysis, risk entity detection, and clinical note processing, enabling automated research and compliance workflows. Applied time-series forecasting models including ARIMA, Prophet, LSTM, and Neural Prophet for financial trend prediction, demand forecasting, and risk assessment in lending, trading, and retail operations. Built comprehensive MLOps and CI/CD pipelines using MLflow, Azure DevOps, Docker, and Kubernetes, ensuring reproducibility, monitoring, and governance in high-stakes ML deployments across regulated industries. Engineered feature engineering and ETL pipelines with PySpark, pandas, Azure Data Factory, AWS Glue, and Delta Lake, delivering enriched, high-quality datasets that significantly improved model performance and business outcomes. Implemented responsible AI practices using SHAP, LIME, Azure AI Content Safety, and fairness metrics to ensure compliance, explainability, and transparency across HIPAA-regulated healthcare and financial services environments. Designed predictive models for fraud detection, claims management, and patient stratification that achieved measurable results including 30% reduction in false claims, 23% improvement in campaign targeting, and 18% enhancement in fraud detection accuracy. Applied knowledge graph models using Neo4j and NetworkX to map complex relationships among customers, transactions, and financial assets for advanced fraud detection, compliance analytics, and risk assessment. Leveraged self-supervised learning and contrastive learning techniques for pretraining on proprietary financial and healthcare text corpora, improving domain adaptation and embedding quality for downstream tasks. Built synthetic data generation workflows using Gretel.ai and internal libraries to augment imbalanced datasets while maintaining data privacy, security compliance, and regulatory adherence in sensitive industries. Created real-time monitoring dashboards using Power BI, Tableau, Azure Monitor, Prometheus, and QuickSight for tracking model drift, prediction quality, system anomalies, and business KPIs across 1,800+ retail locations. Led end-to-end data pipeline development processing high-volume transaction data, healthcare records, and telecom metrics, utilizing distributed computing frameworks and cloud-native architectures for scalable analytics. Delivered cross-functional leadership by mentoring junior engineers, conducting knowledge-sharing sessions on MLOps, generative AI, Azure AI Studio, and foundation model fine-tuning, while translating complex ML models into actionable business insights for stakeholders. Proven track record in statistical modeling, A/B testing, cohort analysis, and demand forecasting that directly contributed to revenue optimization, operational efficiency improvements, and strategic business decision-making across multiple industry verticals. Technical Skills Cloud Platforms Microsoft Azure (Azure ML, Azure Databricks, Azure Synapse, Azure Cognitive Search, Azure OpenAI), AWS (Redshift, S3, Lambda, Athena, Glue, QuickSight) AI/ML Frameworks PyTorch, TensorFlow, scikit-learn, Keras, Cursor AI, AutoGe, Hugging Face Transformers, LangChain, Semantic Kernel MLOps & DevOps MLflow, Azure DevOps, CI/CD Pipelines, Docker, Kubernetes, Prometheus, Azure Monitor NLP & LLMs BERT, GPT, Azure OpenAI, RAG (FAISS, Vector DBs), Text Summarization, Sentiment Analysis, Entity Recognition Time-Series Forecasting ARIMA, Prophet, LSTM, Neural Prophet Big Data & ETL PySpark, pandas, Azure Data Factory, AWS Glue, SQL, Delta Lake Vector & Graph DBs FAISS, Azure Cognitive Search, Neo4j, NetworkX Responsible AI SHAP, LIME, Fairness Metrics, Azure AI Content Safety Data Visualization Power BI, QuickSight, Matplotlib, Seaborn Synthetic Data & Privacy Gretel.ai, Differential Privacy, Data Augmentation Programming Python, SQL, PySpark, Bash Knowledge Graphs Neo4j, NetworkX, Graph Embeddings Self-Supervised Learning Contrastive Learning, Pretraining (Financial Text) Educational Details Bachelors in computer science at Lovely Professional university Aug 2007 to Jun 2011 Masters in computer science at University of Central Missouri Aug 2011 to Dec 2012 Certifications Microsoft Certified: Azure AI Engineer Associate - 2023 AWS Certified Machine Learning Specialty 2021 Work Experience Client: Jefferies Financial Group Inc, New York, NY May 2024 - Present Role: Senior AI/ML Engineer Responsibilities: Designed and implemented AI/ML solutions on Microsoft Azure, leveraging services like Azure Machine Learning, Azure Databricks, and Azure Synapse Analytics for scalable, secure model training and deployment. Led the development of end-to-end ML pipelines using Python, Azure ML SDK, and MLflow, enabling reproducibility, CI/CD integration, and operational model governance. Developed and deployed advanced NLP models for financial document summarization, sentiment analysis, and risk entity detection using transformer-based architectures such as BERT and Azure OpenAI Service. Integrated vector databases like Azure Cognitive Search and FAISS to power retrieval-augmented generation (RAG) pipelines for LLM-based search over internal knowledge bases. Orchestrated LangChain and Semantic Kernel components to build LLM-powered agents for automating financial research and compliance workflows, improving analyst productivity. Built custom RAG pipelines with grounding and context injection strategies to reduce hallucinations and ensure factual accuracy in generative AI outputs. Leveraged Cursor AI to prototype and deploy agentic workflows involving task delegation, retrieval orchestration, and document summarization with integrated validation loops. Built multi-agent systems using LangChain, LangGraph, and Cursor AI for intelligent automation in financial research, streamlining compliance reporting and earnings analysis. Applied time-series forecasting models including ARIMA, Prophet, and LSTM, to predict multi-asset financial trends and perform risk forecasting for lending and trading use cases. Enhanced LLM and NLP output quality using prompt tuning, PEFT (parameter-efficient fine-tuning), and reinforcement learning with human feedback (RLHF) for domain alignment. Utilized Azure Databricks Delta Lake and Unity Catalog for unified governance, lineage tracking, and secure access to financial data lakes across teams and business units. Implemented feature engineering pipelines using PySpark, pandas, and Azure Data Factory, improving model performance through advanced feature transformations and temporal joins. Applied knowledge graph models using tools like Neo4j and NetworkX to map relationships across customers, transactions, and financial instruments for fraud detection and compliance analytics. Integrated responsible AI tooling including Azure AI Content Safety, SHAP, and LIME to ensure model transparency, fairness, and compliance with financial regulations. Designed and deployed LLM-based systems for financial document parsing, earnings summary generation, and real-time investor sentiment extraction via Azure OpenAI. Built synthetic data generation workflows using tools like Gretel.ai and internal libraries to augment imbalanced datasets and protect sensitive financial information. Developed custom monitoring dashboards using Power BI, Azure Application Insights, and Prometheus, to visualize model drift, prediction quality, and data anomalies in production. Led research into self-supervised learning and contrastive learning techniques for embedding and pretraining on proprietary financial text corpora. Implemented data mesh principles and domain-oriented ownership for scalable ML architecture across global financial data domains, fostering decentralized innovation. Mentored junior engineers and led internal sessions on MLOps, generative AI, Azure AI Studio, and foundation model adaptation, establishing best practices for production-scale AI in finance. Client: HCA Healthcare Inc, Nashville, TN Nov 2022 April 2024 Role: AI/ML Engineer Responsibilities: Designed and deployed scalable machine learning models to automate risk prediction and patient stratification within healthcare plans, improving proactive care delivery. Collaborated with cross-functional healthcare and engineering teams to translate clinical goals into data-driven solutions using Azure Machine Learning Studio and Azure Synapse Analytics. Developed robust data pipelines and orchestrated ETL workflows using Azure Data Factory to ensure accurate ingestion and transformation of large-scale healthcare data. Engineered predictive models for claims fraud detection, leading to a 30% reduction in false claims through advanced supervised learning techniques. Built and maintained CI/CD pipelines for ML model deployment using Azure DevOps, improving delivery speed and ensuring reproducibility across staging and production environments. Utilized PySpark, SQL, and Azure Databricks for distributed data processing and real-time analytics in population health management systems. Created explainable AI solutions using SHAP and LIME, enabling compliance with healthcare regulations and enhancing model transparency for clinicians. Integrated external social determinant datasets with HCA data on Azure Data Lake to enhance patient outcome modelling through feature engineering and data fusion. Implemented model monitoring dashboards using Power BI and Azure Monitor, allowing continuous tracking of ML performance and drift in real-world settings. Conducted hyperparameter tuning and model optimization leveraging Azure AutoML for accelerated experimentation and model selection. Ensured all AI solutions complied with HIPAA standards, embedding privacy, security, and governance controls into the ML lifecycle. Supported the development of NLP models to analyze unstructured clinical notes, enabling structured data extraction and classification via BERT and spaCy frameworks. Led efforts in data labelling strategies, annotation workflows, and active learning, accelerating model training for medical image and document classification. Participated in design reviews and code audits to enforce MLOps best practices, including containerization using Docker and orchestration with Kubernetes on Azure AKS. Explored early agentic designs using LangChain and Cursor AI for automating clinical document parsing, note classification, and physician Q&A support. Collaborated with stakeholders to define key metrics and success criteria for AI initiatives, aligning ML outputs with the goals of value-based healthcare. Applied transfer learning and deep learning models to diagnostic imaging use cases, improving anomaly detection accuracy across large datasets. Mentored junior engineers and data scientists on Azure ML workflows, code versioning using Git, and experiment tracking with MLflow. Delivered impactful presentations to both technical and non-technical audiences, demonstrating the business value and clinical implications of AI projects. Client: Target Corp, Minneapolis, MN Jan 2019 Oct 2022 Role: Senior Data Analyst Responsibilities: Led the design and implementation of scalable data pipelines on AWS, utilizing Amazon Redshift, S3, and AWS Glue, to process and store high-volume retail transaction data. Delivered actionable insights by developing complex SQL queries and stored procedures to analyze customer behavior, product trends, and seasonal performance across Target's retail network of 1,800+ locations. Created dynamic dashboard solutions using Tableau and Power BI, enabling real-time executive reporting for inventory optimization and supply chain efficiency. Collaborated with cross-functional teams including marketing, merchandising, and e-commerce to align business goals with data strategies, improving campaign targeting by 23%. Executed deep-dive analysis on sales funnels, conversion rates, and clickstream data using Python (pandas, NumPy) to identify friction points in the customer journey. Managed the end-to-end lifecycle of data modeling projects, ensuring consistency across dimensions and metrics using dbt and Snowflake, enhancing data integrity across teams. Applied advanced statistical methods and regression models to predict demand fluctuations and optimize pricing strategies, contributing to a 12% revenue uplift in 2021. Automated recurring reporting processes with Python scripting and AWS Lambda, reducing manual workload by 40% and improving delivery accuracy. Utilized Amazon QuickSight for visual analytics on retail KPIs, integrating with Athena and Glue Data Catalog for seamless data querying and exploration. Led data governance initiatives, including data quality checks, lineage documentation, and access control using AWS Lake Formation, ensuring compliance with internal and regulatory standards. Integrated external market datasets (e.g., Nielsen, social media sentiment) to enrich internal datasets and support competitive analysis and localized merchandising efforts. Developed forecasting models using Prophet and ARIMA to assist in inventory planning and reduce overstock scenarios across warehouse hubs. Delivered training sessions and onboarding guides for junior analysts on AWS tools, data interpretation, and visualization best practices. Collaborated with DevOps to integrate data monitoring solutions using CloudWatch, improving system reliability and proactive issue detection. Conducted cohort and churn analysis using SQL and Python, identifying high-value customer segments and helping marketing focus on retention campaigns. Implemented best practices in version control, CI/CD pipelines, and analytics code management using Git, AWS CodeCommit, and Jenkins. Acted as a key liaison between business stakeholders and technical teams, translating analytical findings into business decisions that directly impacted revenue, customer engagement, and operational efficiency. Client: Allstate, Northbrook, IL Sept 2015 Dec 2018 Role: Data Analyst Responsibilities: Conducted comprehensive data analysis on policyholder behavior, claim trends, and underwriting data to support strategic insurance pricing decisions and risk assessments. Utilized AWS services (S3, Redshift, Athena, and Glue) to manage and process large insurance data sets securely and efficiently within a cloud-based infrastructure. Developed and maintained interactive dashboards using Tableau and Power BI to track KPIs, claim volumes, and customer retention trends, increasing stakeholder visibility into performance metrics. Designed and implemented ETL pipelines leveraging AWS Glue and Python to automate data ingestion from multiple sources including internal CRM and third-party actuarial datasets. Collaborated with actuarial and underwriting teams to perform predictive modeling using Python (pandas, scikit-learn) for churn prediction and fraud detection in claims processing. Optimized SQL queries on Amazon Redshift to deliver near real-time reporting and improved system performance by 30% in monthly executive reporting. Participated in data governance initiatives to ensure data quality, accuracy, and compliance with insurance regulations including HIPAA and state-level privacy laws. Conducted detailed cohort analyses to identify customer behavior patterns and delivered actionable insights that directly improved cross-sell campaign effectiveness. Led efforts in integrating and standardizing structured and unstructured data across policy, claims, and customer interaction channels, using Python and AWS Lambda functions. Applied advanced statistical methods and machine learning techniques to support pricing optimization and dynamic segmentation of customers. Coordinated with business stakeholders to gather requirements and translate them into analytical models and visual solutions that directly influenced operational decisions. Created data dictionaries and maintained metadata repositories to ensure data transparency and traceability across functional teams and tools. Provided ad hoc reporting and strategic insights to support sales, marketing, and claims departments in aligning performance with business goals. Implemented data anomaly detection scripts in Python and SQL to flag outliers in claim submissions, improving fraud detection by 18%. Collaborated with DevOps and cloud engineering teams to ensure secure and scalable deployment of analytics workflows using AWS IAM, EC2, and CloudWatch. Regularly evaluated new market analytics tools and insurance industry trends to enhance analytical capabilities and align with evolving data science practices. Client: Ooma Inc, Sunnyvale, CA Feb 2013 Aug 2015 Role: Data Analyst Responsibilities: Collaborated with cross-functional teams to analyze large-scale telecom data, improving decision-making processes and increasing operational efficiency by over 20%. Developed and maintained automated dashboards and visual reports using Tableau and Power BI, enabling senior leadership to monitor KPIs in real-time. Executed end-to-end data analysis projects by extracting data from SQL databases, transforming it through ETL pipelines, and performing statistical evaluation. Conducted deep analysis on customer behavior and churn patterns using Python (pandas, NumPy) and R, leading to actionable retention strategies. Worked closely with engineering and marketing teams to align data findings with telecom industry trends, enhancing product targeting and customer segmentation. Implemented data validation and cleansing routines to ensure data integrity and consistency across systems, improving report accuracy by 30%. Led efforts in market segmentation analysis, providing insight into user demographics and usage patterns which supported strategic pricing initiatives. Created predictive models using machine learning techniques to forecast call drop rates and optimize network resource allocation. Processed large datasets using SQL Server and MySQL, optimizing queries for analytical performance. Worked with Oracle databases and basic ETL processes to support telecom analytics. Assisted in migration projects by validating datasets post-transfer, ensuring smooth transition and compliance with industry standards. Supported monthly and quarterly business reviews with trend analysis, highlighting growth opportunities and operational inefficiencies in telecom services. Developed A/B testing frameworks to assess the impact of new telecom features, helping guide product development with data-driven insights. Provided stakeholder training on interpreting data reports and visualizations, promoting a data-driven culture across departments. Analyzed call quality and VoIP metrics, identifying bottlenecks and recommending improvements that enhanced service reliability and customer satisfaction. Maintained documentation for data models, processes, and business logic to ensure transparency and enable future scalability. Keywords: continuous integration continuous deployment artificial intelligence machine learning business intelligence sthree active directory rlang information technology trade national California Illinois Minnesota New York Tennessee