Challenge
One of the leading energy companies in the region was facing a significant increase in internal inquiries related to production operations, supply chain logistics, and well management. These questions were being handled by highly specialized technical teams, creating bottlenecks and increasing workload on personnel.
The company was seeking a solution that would:
- Automate the resolution of technical inquiries without compromising response quality.
- Unify multiple knowledge sources into a single point of contact.
- Integrate with the client’s existing collaboration tools.
Solution
CloudHesive implemented a solution powered by Amazon Bedrock, developing three specialized GenAI agents that work together to provide a seamless and accurate conversational experience through a single interface.
Specialized Agents:
- Agent 1 – Roles: Answers questions about processes, permissions, organizational structure, and job roles.
- Agent 2 – Wells: Delivers detailed information on wells, production, and operations.
- Agent 3 – Supply Chain: Supports logistics, orders, materials, and delivery inquiries.
These agents operate as one unified assistant, capable of responding based on the domain of the question without requiring users to distinguish between them.
Microsoft Teams Integration
To maximize adoption and ease of use, the assistant was directly integrated into Microsoft Teams, allowing employees to interact with GenAI without leaving their familiar collaboration environment.
Expected Results
- Reduced operational load on key technical areas.
- Response times decreased from 10+ minutes to under 1 minute.
- Over 60% of inquiries resolved without human intervention.
- Immediate scalability to include new agents or workflows.
- High adoption rate due to native Teams integration.
Technology Stack
- Amazon Bedrock (LLM Foundation Models)
- Amazon Lex
- AWS Lambda
- Amazon API Gateway
- Microsoft Teams (via custom webhook/connector)
- IAM, VPC, Secrets Manager
