Skip to main content

Pharia AI Applications

This document provides a brief introduction to Pharia AI Applications, including their motivation, architecture, components, and how to get started. It is intended for developers, architects, and technical stakeholders who want to understand how to build and deploy AI-powered applications using the Pharia AI platform.

Motivation for Pharia AI Applications​

Building production-ready AI applications presents significant challenges that shift focus away from solving real business problems:

  • Infrastructure setup and configuration can be time-consuming, delaying the start of development.
  • Prototyping is hindered by technical overhead related to data pipelines, model deployments, and APIs.
  • AI application development tools are dispersed across various open-source solutions, requiring different setups and integration.
  • Prototypes often face obstacles when transitioning to production.
  • Available interfaces are typically limited to generic chat interfaces.

Teams aiming to quickly develop and deploy AI solutions often get bogged down by infrastructure setup, technical overhead, tool integration, and moving prototypes to production, instead of focusing on generating business value.

Pharia AI + Pharia AI Applications address these challenges by providing:

  • A complete suite of production-grade infrastructure and tools.
  • The ability to rapidly build production-grade prototypes that can scale seamlessly.
  • Built-in best practices for reliability and scalability.
  • Tooling for tracing, debugging, and evaluation.
  • A unified but customizable interface for interacting with AI applications.
  • A platform to build reusable AI logic and out-of-the-box AI capabilities.

With Pharia AI Applications, teams can focus on validating their use cases while building on a foundation that is production-ready from inception, significantly reducing time-to-market and development costs.

What is a Pharia AI Application?​

A Pharia AI Application is an end-to-end solution to a business problem that is production-ready from day one. It is developed using the Pharia AI Application framework and leverages the Pharia AI infrastructure. A typical Pharia AI Application is composed of:

  • An Application UI
  • An Application Service

A Pharia AI Application leverages:

  • One or more pre-built or custom Pharia AI Skills that offer reusable AI functionalities.
  • Production-grade infrastructure the provides necessary functionality for model serving and inference, data management, vector search, tracing and evaluations.

This framework enables development teams to rapidly build and deploy AI applications by focusing on their unique business requirements while leveraging Pharia's infrastructure components and reusable software components. Teams can create completely customized solutions while significantly reducing time-to-market and development costs through built-in production readiness.

Architecture​

Pharia Application Architecture

Components of a Pharia AI Application​

Application UI​

The Application UI serves as the primary interface for user interactions within a Pharia AI Application. It is responsible for capturing user inputs, facilitating communication with the Application service, and defining the user experience. The Application UI can currently be built using Vue.js and is setup as a microfrontend application that can be rendered in Pharia Assistant.

Application Service​

The Application Service acts as the backend engine for a Pharia AI Application. It orchestrates data flows, maintains the application state, and provides APIs essential for the Application UI. By leveraging Pharia AI Skills, it addresses specific AI use cases efficiently. Additionally, it manages critical data storage needs, including user sessions and application-related data.

PhariaAI Skill​

A PhariaAI Skill is a specialized, modular function designed to solve a specific AI challenge using a Model. Built and served using Pharia Kernel, Skills form the foundational building blocks of a Pharia AI Application. Their modular design encourages reuse across multiple applications, promoting efficiency and consistency in solving diverse AI problems. The development of a Skill focuses on the AI logic while utilising Pharia Kernel to handle interaction with the Inference API, Scaling, Deployment and Tracing/Evaluations via integration with Pharia Studio.

Components of Pharia AI for building an Application​

Getting Started​

To begin building your own Pharia AI Application, refer to our comprehensive guides:

These guides will walk you through the setup process, basic concepts, and best practices for developing Applications with Pharia AI.