Built centralized MCP servers to standardize tool invocation, context exchange, and permissioned data access between agents, enabling secure, versioned, and auditable AI interactions across the platform

Implemented Agent-to-Agent (A2A) communication patterns, allowing specialized agents (retrieval, reasoning, validation, action) to collaborate via MCP-defined contracts within LangGraph agent graphs

Developed stateful LangGraph agent workflows with task decomposition, conditional branching, retries, and fallback handling, closely aligned with Copilot Studio and Power Automate orchestration patterns

Integrated LangGraph agents with Azure OpenAI Service, Azure AI Search vector indexes, and enterprise systems through custom APIs and HTTP-based connectors, enabling secure enterprise automation scenarios

Developed secure backend services integrating RBAC, MCP-enabled agents, and A2A workflows with a React frontend using Azure-hosted microservices, enforcing Microsoft Entra ID (AAD) authentication

Engineered advanced chunking, embedding, and indexing pipelines feeding LangGraph retrieval nodes, supporting PDFs, Word, Excel, PPT, and SQL data for accurate and governed grounding

Conducted stakeholder intake workshops to translate enterprise requirements into LangGraph agent graphs, MCP tool definitions, and Azure-native RAG designs

Optimized model performance using Azure OpenAI–compatible fine-tuning, LoRA, and PEFT techniques to improve domain accuracy while reducing latency and compute cost

Conduct hands-on workshops and lab sessions, guiding students through the implementation of generative AI models using frameworks like Tensorflow, PyTorch and Hugging Face Transformers.

Support research projects that explore innovative applications of LLMs in areas such as financial forecasting, supply chain optimization, and customer behavior analysis.

Provide one-on-one mentoring to students, helping them design and implement AI-driven solutions for their capstone projects.

Collaborate with faculty to integrate real-world case studies that demonstrate the impact of generative AI on business strategy and decision-making processes.

Assist in the evaluation and grading of student projects, providing constructive feedback on both technical implementation and business applicability.

Contribute to ongoing research in the field of AI applications in business, co-authoring papers and presenting findings at academic conferences.

Implemented an NLP-driven chatbot using BERT and GPT-3, capable of handling routine customer inquiries and providing personalized treatment information and content for customer success agents.

Created a machine learning pipeline for cross-selling recommendations, utilizing collaborative filtering and content-based approaches, resulting in increase in additional product sales.

Designed and deployed an SEO optimization system using NLP techniques to analyze competitor content, identify keyword opportunities, and generate SEO-friendly content suggestions, leading to improvement in organic search rankings.

Developed a CLTV prediction model using gradient boosting algorithms, incorporating treatment adherence data, purchase history, and engagement metrics to forecast and optimize customer value.

Implemented a real-time data processing pipeline using Apache Kafka to ingest and analyze customer feedback, treatment progress data, and website interactions for dynamic personalization.

Created a full-stack web application using React for the frontend and FastAPI for the backend, providing an intuitive interface for customers support and a comprehensive dashboard for dental professionals.

Utilized AWS SageMaker to train, deploy, and manage machine learning models at scale, including automated retraining pipelines to continuously improve model accuracy with new data.

Designed and implemented a data lake architecture using AWS S3 and AWS Glue, consolidating data from various sources including CRM systems, treatment devices, and customer interactions.

Implemented natural language processing (NLP) techniques using BERT and transformers to analyze financial reports, news articles, and social media sentiment, enhancing the company's ability to gauge market sentiment and identify emerging trends.

Created a real-time anomaly detection system using Azure Stream Analytics and custom machine learning models to identify unusual market behavior and potential fraud, reducing false positives.

Designed and implemented a recommendation engine for financial products using collaborative filtering and content-based approaches, increasing cross-selling opportunities.

Developed time series forecasting models using ARIMA, Prophet, and LSTM networks to predict stock prices and market volatility, achieving improvement in forecast accuracy compared to traditional methods.

Utilized Azure Databricks to build scalable data processing and feature engineering pipelines, optimizing the workflow for training and deploying machine learning models on large financial datasets.

Implemented automated machine learning (AutoML) pipelines using Azure Machine Learning, enabling rapid prototyping and model selection for various predictive tasks across different business units.

Collaborated with data engineers to design and optimize data schemas in Azure Synapse Analytics, ensuring efficient storage and retrieval of features for machine learning models.

Developed custom Python libraries and reusable code modules for common data science tasks, improving team productivity and ensuring consistency in model development across projects.

Implemented a real-time anomaly detection system using Azure Stream Analytics and custom machine learning models to identify unusual patterns in IoT sensor data, enabling proactive maintenance and improving equipment reliability.

Created a predictive analytics pipeline using Azure Databricks and PySpark to forecast project timelines and resource requirements, improving project planning accuracy and optimizing resource allocation.

Designed and implemented a natural language processing (NLP) solution using BERT and Azure Cognitive Services to analyze safety reports and identify potential hazards, enhancing site safety protocols and reducing incident rates.

Developed a computer vision model using TensorFlow and Azure Custom Vision to automate progress monitoring of construction sites through image analysis, increasing the accuracy of project status reporting.

Implemented a recommendation system for optimal equipment usage and resource allocation using collaborative filtering techniques, leading to increase in overall equipment efficiency.

Utilized Azure Synapse Analytics to build a centralized data warehouse, integrating data from various sources including IoT sensors, project management software, and financial systems for comprehensive analysis and reporting.

Created interactive Power BI dashboards to visualize key performance indicators, machine learning model outputs, and predictive insights, facilitating data-driven decision-making for project managers and executives.

Developed time series forecasting models using ARIMA, Prophet, and LSTM networks to predict material costs and optimize procurement strategies, resulting in reduction in overall procurement costs.