The client is a global aviation maintenance and support provider. The company operates across the United States, the Bahamas, and the United Kingdom, managing 48 line‑maintenance stations that support commercial, military, and cargo aviation.
With a strong focus on reliability and speed, the client plays a critical role in keeping fleets running efficiently across multiple regions and operational environments.
This client faced the challenge of having a powerful, intelligent AI-driven support tool for both external users and internal teams, enabling centralized access to company information, supporting key processes like part requests and purchase orders, and providing a scalable foundation for future enhancements.
A second challenge was the client’s information was distributed across multiple institutional websites that were difficult to navigate, creating a strong need for a unified system capable of providing fast, accurate answers in a single place.
The initiative also required close coordination across several internal teams. The AI-Assistant needed to integrate with an existing database managed by a dedicated team, while infrastructure and IT teams were responsible for enabling secure operations and ensuring backend services ran reliably. This meant establishing communication channels, aligning efforts across groups, and continuously monitoring progress to prevent bottlenecks.
To help address these challenges, the team built a comprehensive AI‑assistant designed for both internal and external users. The solution provides instant answers about the company and supports operational processes such as part requests and purchase order inquiries. At its core, a Python API powered by LangGraph orchestrates the system, intelligently routing user queries to the appropriate actions or data sources.
For transactional needs such as retrieving parts information, the assistant integrates directly with a SQL database. For organizational knowledge queries, the team implemented a Retrieval‑Augmented Generation (RAG) service connected to a collection of documents stored in an S3 bucket. These documents, which come from both structured files and scraped institutional websites, are indexed as vectors to support accurate, context‑aware responses.
To streamline content updates, the system includes a daily AWS Lambda process that automatically transfers documents uploaded to SharePoint into S3, ensuring the AI-assistant always has access to the latest information. The solution also incorporates user authentication with domain‑based access control, a canvas for RFQs, downloadable data capabilities, and an admin panel for managing user roles.
Four years into their partnership, this aircraft company not only depends on Hikru for comprehensive IT support but also utilizes their software & AI developers for crucial projects. This collaboration has resulted in notable labor cost savings and consistent, high-quality project delivery. As the company grows, it relies on Hikru's capability to scale and adapt to its changing needs.
