How Mission Empowers Netfor's Next-Generation Agentic AI-Powered IVR Solution

Netfor, a people-first Business Process Outsourcing (BPO) company, worked together with Mission to develop their most ambitious project: a next-generation agentic AI-powered Interactive Voice Response (IVR) system. Through Mission's expertise in AWS technologies and agentic AI architecture, Netfor is creating an innovative autonomous solution that intelligently leverages its extensive knowledge base to enhance customer service capabilities while maintaining a human touch.

Netfor is a leading BPO provideroffering tailored managed services including field services, first-level tech support, and call center support. Their customer-centric approach focuses on delivering efficient, personalized solutions across government contracts and retail sectors, serving approximately 30-40 diverse clients.

Netfor recognized the need to innovate its customer service capabilities through agentic AI integration. With over 16,000 active knowledge articles developed over three decades, they sought to leverage this valuable resource through an autonomous AI agent that could make intelligent decisions about how and when to access information, creating a more efficient and cost-effective service delivery model.

Netfor chose Mission as its AWS partner for their deep technical expertise in agentic AI architecture and collaborative approach to solution development. Rather than a traditional vendor-customer relationship, Mission offered a true partnership model, working alongside Netfor to design and implement a sophisticated agentic system that could autonomously handle complex decision-making workflows and tool orchestration. As David Cady, Director of Data Operations at Netfor, emphasized, "Mission has been extremely wonderful to work with. Their depth of knowledge is incredible. Their talent is incredible. Their ability to work with you, understand your problems, and help you develop a solution that is very specific to what you are trying to do. It's been a great partnership that we've had with them."

AWS's comprehensive suite of AI and machine learning services provided the robust foundation needed for Netfor's agentic IVR solution. Amazon Bedrock's foundation models enabled the autonomous decision-making capabilities at the heart of the agent, while Amazon Connect provided unlimited concurrent call handling. The platform's integration of AWS Lambda for agent orchestration, Amazon OpenSearch for knowledge retrieval, and other managed services created a scalable, secure environment for deploying production agentic AI workloads with optimized security posture and cost efficiency.

Mission worked closely with Netfor to design and implement a sophisticated agentic AI system that autonomously manages customer interactions through intelligent tool selection and decision-making. The solution centers on an agent orchestrator deployed in AWS Lambda that coordinates the agent's autonomous operations, leveraging Amazon Bedrock foundation models to power the decision-making engine at the heart of the system.

The agent independently evaluates each customer query to determine the appropriate response strategy. Rather than following rigid rules, the agentic system intelligently decides whether a question requires a direct response, a RAG-based knowledge base lookup, conversation history retrieval, or escalation to human agents. Mission implemented multiple tools that the agent can autonomously call, including knowledge base search through Amazon OpenSearch for retrieval-augmented generation from Netfor's 16,000 knowledge articles, conversation transcript access for context-aware responses, automated ticket creation, and intelligent handoff to human representatives for complex queries.

Mission carefully selected Amazon Bedrock models to optimize performance. The team chose Amazon Nova Micro for primary agent interactions due to its exceptional speed, crucial for maintaining low latency in telephony applications and ensuring users receive responses promptly. Additionally, Mission utilized Amazon Titan v2 for vector embeddings when the agent performs RAG-based knowledge searches. The Lambda-based orchestrator manages the decision workflow, parsing the foundation model's JSON responses to determine which tools to invoke and in what sequence, creating a truly autonomous system where the LLM makes core decisions about whether RAG lookup is needed. As Cady noted, "Mission's role in this project has been instrumental. Not only do they help us design basic architecture, but they have really worked with us as a partner...it very much feels like we are all on the same side."

Mission implemented comprehensive security controls for the agentic system, including prompt guardrails and backend logic filters with company ID-based restrictions to ensure users only received responses derived from authorized documents. The team adhered to the principle of least privilege, granting the agent role permissions only to necessary knowledge base documents. Through iterative testing with the Netfor team, including attempts at prompt injection attacks via voice, Mission validated that responses were appropriate and access was properly controlled.

The solution is customizable for each customer's needs, allowing Netfor to tailor the agent's knowledge base based on individual campaign requirements, whether for retail stores or government contracts.

The partnership with Mission has positioned Netfor at the forefront of agentic AI innovation in customer service. The autonomous IVR system promises significant benefits for Netfor's clients, including substantial cost savings through efficient autonomous call handling and improved first-call resolution rates. The agent operates 24/7, handling straightforward queries without human intervention while intelligently routing more complex issues to live representatives. Even when calls require human intervention, the system reduces handling times by gathering and passing along relevant information to agents, as the agentic system has already pre-processed the context.

The solution's ability to handle unlimited concurrent calls through Amazon Connect ensures scalability during peak periods, addressing a critical operational challenge. As Cady notes, "Using Amazon Connect, we can handle as many phone calls at once as required. When we suddenly get a thousand calls in an hour, we'll have zero problems handling those -- something that would be much more challenging with live agents."

Most importantly, the partnership with Mission has enabled Netfor to pursue cutting-edge agentic AI innovation while maintaining its commitment to quality service. In Cady's words, "If we weren't using Mission to help us develop this product, it would be extremely challenging. We don't have the level of expertise that we would need, so it would be very difficult to make sure that we didn't have technical issues."

With this cutting-edge agentic AI solution, Netfor is positioned to become a leader in their industry. The autonomous agent's ability to make intelligent decisions about tool selection, combined with its seamless integration with human agents when needed, demonstrates the power of properly architected agentic AI systems. As they continue to develop and refine their AI-powered solution, they're excited about the possibilities for revolutionizing customer service delivery while maintaining their commitment to innovation and solving industry problems.