Innovative IT Solutions for Your Business

The Challenge: A company wanted to modernize its customer service by providing a seamless, voice-driven experience for order processing and payments, reducing the need for human agents and improving customer satisfaction. The solution needed to be robust and integrate with existing phone systems.

The Solution: We architected and led the deployment of a real-time, AI-powered voice assistant. The solution utilized an architecture of microservices, Twilio for phone integration, Azure AI Search for data retrieval, and Large Language Models (LLMs) to understand and process customer requests. I deployed the entire solution on Azure, ensuring a scalable and reliable platform for real-time customer interactions.

The Impact: The new voice agent provides an intuitive, hands-free experience for customers, automating key service tasks and enabling the company to handle a higher volume of calls more efficiently.

The Challenge: A business needed a flexible way to create and deploy custom AI agents that could be tailored to different workflows and easily integrated into their company’s culture. The goal was to accelerate AI adoption without requiring extensive technical expertise for each new agent.

The Solution: We architected and directed the implementation of "AAPI (Agent APIs)" a scalable agentic platform on Google Cloud Platform (GCP). The platform was built on Kubernetes for scalability, Cassandra with vector DB capabilities for RAG features, and leveraged Llama 3.1 70B for powerful language processing. This design provides a rich set of primitives for building AI agents that can be consumed as APIs (Like Open AI Agents API), abstracting the creation process and allowing companies to quickly deploy custom "AI Employees" that align with their specific needs.

The Impact: AAPI provides a robust and repeatable framework for creating tailored AI agents, significantly reducing the time and resources needed for AI development and accelerating the company’s digital transformation.

The Challenge: A sales organization was struggling with lead retention and needed a way to provide their sales teams with real-time, contextual information during live meetings. They required a solution that could transcribe conversations and provide instant insights from existing company documents.

The Solution: We implemented a real-time AI-powered sales assistant platform on GCP. The solution included a Web Portal and a local agent that used a secure WebSocket connection. The agent monitored live calls, sent real-time audio transcriptions to Vertex AI models, and used a PostgreSQL as a vector database for document similarity search. The platform also leveraged the spaCy library for entity and topic extraction, providing sales users with immediate, data-driven suggestions during their calls.

The Impact: The platform increased lead retention by empowering sales teams with contextual document search, intelligent suggestions, and conversational AI support during live meetings, giving them a significant competitive advantage.

The Challenge: Project managers needed a solution to streamline their project methodology and task management within the company's Microsoft 365 ecosystem. The goal was to enable PMs to create tasks and align workflows using natural language.

The Solution: We engineered and implemented a project management agent assistant within the Microsoft 365 ecosystem. The solution was deployed on Azure using Azure Container Apps, Azure AI Search for document retrieval, and Bot Services for interaction within MS Teams. We designed the architecture and built the Infrastructure as Code (IaC) and CI/CD pipelines to ensure the solution's robustness and scalability, allowing PMs to use natural language to create tasks in Azure DevOps and collaborate more effectively.

The Impact: The PM Agent Assistant streamlined task creation and workflow alignment, improving collaboration and ensuring project methodology was consistently followed, all within the familiar Microsoft 365 environment.