Client
A global media and information conglomerate, with a diverse portfolio spanning magazines, television, and digital platforms. The organization has a strong presence in the publishing industry, focusing on delivering high-quality content across multiple channels.
Business Challenge
The client wanted to assess their readiness for Generative AI adoption and ensure seamless integration of Conversational AI technology for L1 support.
Key requirements included:
- AI readiness assessment: Assessing existing support workflows, ticketing systems, and AI adoption levels to determine feasibility.
- Infrastructure & data assessment: Evaluating data infrastructure and automation potential to align AI capabilities with business needs.
- Technology selection: Identifying suitable AI models (Claude Sonnet for entity extraction, Amazon Titan for embeddings) and AWS services (Lambda, Pinecone, SES) to support intelligent automation.
- Optimizing AI-driven support: Implementing RAG-based knowledge retrieval, ticket classification, and real-time translation while ensuring effective human escalation.
Solution
Nous adopted a structured three-phase approach:
Maturity assessment – Conducted an AI readiness evaluation of support workflows, data infrastructure, and integration feasibility, identifying key gaps in automation and ticket handling efficiency.
Strategy – Based on the assessment, selected the Claude Sonnet model for entity extraction, and Amazon Titan with Pinecone for vector search, and AWS services (Lambda, SES, etc.) for seamless automation and deployment.
Implementation – Deployed RAG-based knowledge retrieval, automated ticket classification, and real-time translation with human escalation for complex queries.
Key Features:
- Optimized tool selection: Implemented tailored AI solutions to meet business needs.
- Iterative refinement: Used an agile approach to fine-tune processes and enhance results.
- Seamless collaboration: Implemented clear escalation protocols for effective interaction between automated systems and human agents.
- Continuous improvement: Established a regular feedback loop to align with evolving business needs.
Technologies
Business Benefits
- Reduced IT support costs by 15-20% by minimizing manual intervention.
- Lowered backend expenses by 10-15% with scalable AWS services like Lambda and Pinecone.
- Boosted AI utilization in support teams by 35-40% in three months.
- Improved AI-driven insights by 30-40% with optimized tool selection.
- Scaled support system to handle 50% more requests without extra staffing.
Outcome
Reduced average ticket handling time by 30-40% and accelerated resolution by 20-25% through automation, real-time translation, and efficient knowledge retrieval and classification.
