Adanomad Adanomad
Challenge Internship

AI Full Stack Developer Intern Challenge

Coding challenge for Ada Nomad's AI Full Stack Developer Intern position

AI Full Stack Developer Intern Challenge

About Ada Nomad

Ada Nomad is an AI software consulting company founded in March 2023. We develop AceTeam.AI and empower productivity through AI agents and workflows. Our founder, Jason Sun, brings a wealth of experience from top Silicon Valley companies like Apple and Amazon, as well as from self-driving robotics startups.

Challenge Overview

This challenge is designed to align with our intern roles and allow you to showcase the skills in areas relevant to our work. Please choose one of the following tracks based on the interests and expertise:

  1. Web Full Stack Engineer
  2. LLM AI Backend Engineer
  3. Security and Protocol Engineer
  4. Computer Vision Engineer

You have the flexibility to complete this challenge at your own pace. Feel free to use any programming languages, libraries, and frameworks you’re comfortable with, though we recommend using technologies mentioned in our job descriptions when possible.

Please submit the solution before the scheduled interview. During the interview, we will review the work together, and you’ll have the opportunity to modify the code and discuss the approach.

Challenge Tracks

1. Web Full Stack Engineer

Focus: Typescript, React, Next.js

Task: Develop a web application for an AI Agent Dashboard

Requirements:

  • Create a responsive frontend using React and Next.js
  • Implement a backend API using Node.js or Python
  • Design an interface for AI agents to interact with users and other AI agents
  • Implement a simple workflow system for AI agent tasks

Bonus:

  • Integrate with a third-party AI API (e.g., OpenAI, Hugging Face)
  • Implement real-time updates using WebSockets

2. AI LLM Backend Engineer

Focus: RAG (Retrieval-Augmented Generation), Knowledge Graphs

Task: Develop a multi-modal knowledge graph system with improved RAG responses

Requirements:

  • Implement a basic knowledge graph structure
  • Create an API for querying and updating the knowledge graph
  • Develop a RAG system that utilizes the knowledge graph for improved responses
  • Implement at least one AI workflow that enhances an AI agent’s capabilities

Bonus:

  • Incorporate multi-modal data (text, images, etc.) into the knowledge graph
  • Implement a simple visualization of the knowledge graph

3. Security and Protocol Engineer

Focus: Privacy-preserving RAG, Zero-Knowledge Proofs

Task: Develop a privacy-preserving RAG framework with secure agent-to-agent communication

Requirements:

  • Implement a basic RAG system
  • Add a layer of privacy preservation to the RAG system (e.g., homomorphic encryption)
  • Create a simple protocol for secure agent-to-agent communication
  • Implement a basic private information retrieval system

Bonus:

  • Incorporate Zero-Knowledge Proofs into the solution
  • Develop a simple blockchain-inspired protocol for agent-to-workflow interactions

Certainly! I’ll add a fourth track for Computer Vision Engineer and provide the sections under it. Here’s the addition:

4. Computer Vision Engineer

Focus: Image and Video Analysis, Pose Estimation

Task: Develop a BJJ (Brazilian Jiu-Jitsu) Pose Classification System

Requirements:

  • Implement a computer vision model to detect and classify BJJ poses from images or video frames
  • Create a simple web interface to upload images or video clips for analysis
  • Develop an API endpoint that accepts image/video input and returns pose classifications
  • Implement real-time pose classification for video input (if choosing video analysis)

Bonus:

  • Extend the system to track and analyze sequences of poses over time
  • Implement a feature to provide feedback on technique based on the detected poses
  • Create a visualization component to overlay pose estimations on the input images/video

Submission Guidelines:

  • Provide the trained model or instructions on how to use a pre-trained model
  • Include sample images or video clips that demonstrate the system’s capabilities
  • If using a notebook environment like Google Colab, provide the notebook with clear instructions

Resources:

Submission Guidelines

Please provide the following as the submission:

  1. The source code of the solution, as a GitHub repository or email a zip file.
  2. Include a README file with:
    • Instructions for running the application and API
    • A list of technologies and libraries used
    • Any assumptions or design decisions you made
  3. A brief explanation of the approach and how it aligns with the chosen track
  4. A short demo video (2-3 minutes) showcasing the solution’s functionality

Evaluation Criteria

Your submission will be evaluated based on the following criteria:

  • Functionality: Does it work as intended?
  • Code Quality: Is the code well-structured, readable, and maintainable?
  • Innovation: How creative and effective is the approach?
  • Alignment: How well does the solution align with the chosen track and Ada Nomad’s focus areas?
  • Documentation: Is the submission accompanied by clear and comprehensive documentation?

Resources

We look forward to seeing the innovative solutions! If you have any questions or need clarification, please don’t hesitate to reach out. Good luck!