Filter

AI-Powered Workflow Agent in a Sandboxed Environment

Description

Why this is important?

AI Agents enhance large language models (LLMs) by performing user-driven actions, enabling various commercial applications. This is a nascent domain will emerging frameworks such as the model context protocol (MCP) leading to commercial products and services. The Arm architecture, from microcontrollers to servers, will be used to carry out agentic functions and Arm has many initatives to support the AI future. See our website for more details.

Project Summary

Participants must develop an AI-powered agent that automates repetitive and complex workflow tasks in a specific domain, such as software development, e-commerice, or DevOps. The foundational model can be a suitable model of your choice (e.g., OpenAI API) but you must consider the appropriate model for cost, reliability and accessibility. Additionally, you are free to choose the tools for agent functionality, such as LLama-cpp-agent. One stipulatation, is that the LLM and/or agent must run on an Arm-based system, such as a Google Pixel phone or Arm-based server.

The AI agent will be deployed in a sandboxed environment to ensure safety and prevent unintended consequences, including prompt guardrails

Prerequisites

  • Intermediate understanding in an OOP language such as Python (for front-end, if needed).
  • Familiarity using Databases such as PostgreSQL, MongoDB, VectorDB.
  • Access to a LLM (e.g., through an API or on-device LLM)
  • Optional API access to target workflow tools such as Jira, Jenkins etc.

Resources from Arm and our partners

Support Level

This project is designed to be self-serve but comes with opportunity of some community support from Arm Ambassadors, who are part of the Arm Developer program. If you are not already part of our program, click here to join.

Benefits

Standout project contributions will result in preferential internal referrals to Arm Talent Acquisition (with digital badges for CV building). And we are currently discussing with national agencies the potential for funding streams for Arm Developer Labs projects, which would flow to you, not us.

To receive the benefits, you must show us your project through our online form. Please do not include any confidential information in your contribution. Additionally if you are affiliated with an academic institution, please ensure you have the right to share your material.

Previous Submissions

  1. AI to Solve Maths Example Sheets at University of Cambridge. (Finley Stirk, Eliyahu Gluschove-Koppel and Ronak De)

  2. AI that interprets user requests, generates circuit descriptions, creates LTSpice ASC code, and iteratively refines circuit designs using a combination of GPT-based language models, a vision analysis module, and LTSpice simulation. (Gijeong Lee, Bill Leoutsakos)

  3. AI agent to track real-time student engagement and exam performance (Jasper Wang, Sritej Tummuru, Talha Javed)