Challenges/TalentReach/General/Agentic AI Engineer Application Challenge -- Designing an AI-Powered Workflow

    Agentic AI Engineer Application Challenge -- Designing an AI-Powered Workflow

    AI Assistant for Busy Clinics

    AI
    LLM
    Agent Design
    System Architecture
    Healthcare Tech
    API Integration
    User Experience
    Full Stack Development
    Estimated Time:
    10 minutes
    Difficulty:Advanced
    Status:Not started
    Start this challenge

    Create a free account to upload your work. Your progress saves as a draft until you submit.

    What You'll Be Doing

    Context: You've joined a team building a platform used by medical and dental practices. The company wants to ship its first agentic AI feature: an assistant that helps clinic staff handle their daily workload more efficiently.

    Today, staff manually:

    • Read through incoming messages and requests (patients asking for appointments, refills, billing questions)
    • Check practice management / EHR systems for patient info, schedule, and eligibility
    • Decide what to do next: reply, schedule, escalate to a doctor, or ask for more info

    Leadership wants an AI-powered workflow where:

    • Messages are triaged and summarized using AI
    • The system proposes recommended actions (e.g., "offer these 3 time slots," "route to billing," "flag for doctor review")
    • Staff can review, edit, and approve actions in a React-based UI, and then the platform executes them via backend services and APIs

    Record a video walking us through how you would approach designing and building this system. You may also include optional notes or diagrams to support your presentation.

    Challenge Guidelines

    Assessment Format:

    • Video Response: One continuous recording, maximum 10 minutes
    • Written Artifacts: Notes or diagrams used to prepare your video submission
    • Assignment: One challenge per candidate
    • AI Usage Declaration: Required

    Video Requirements (≤ 10 minutes total):

    Please record one continuous video responding to the prompts below.

    Part 1: Personal Context (≈ 2 minutes)

    Talent Reach wants to understand your broader career goals. If this role isn't the perfect fit, we'd love to help point you toward other opportunities.

    Please share in your video some or all of these:

    • Why this role? What excites you about being a Technical Implementation Lead?
    • If not this role, what other roles interest you? (e.g., Solutions Architect, Product Manager)
    • Company environment: Do you prefer startup, growth-stage, or enterprise settings?
    • Culture fit: What matters most to you in a company culture? (e.g., innovation, stability, mentorship, diversity)
    • Career trajectory: Are you focused on individual contributor work or interested in leadership/management paths?
    • Open to contract, permanent, or both?
    • Salary range or hourly rate
    • Remote, hybrid, or in-office preference
    • Travel requirements acceptable?
    • Tech stack
    • Product development vs. infrastructure
    • Greenfield vs. maintenance
    • Preferred methodologies (Agile, Scrum, etc.)
    • Non-negotiables
    • Visa sponsorship needs

    Part 2: Challenge Scenario --- "AI Assistant for Busy Clinics" (≈ 8 minutes)

    Address the three areas outlined in the Challenge Description.

    What You'll Accomplish

    Design full-stack system architecture for AI-powered workflows

    Implement safety guardrails and human-in-the-loop controls

    Build transparent and explainable AI interactions

    Balance scalability with simplicity in first iterations

    Apply AI-assisted development tools thoughtfully

    How Your Work Will Be Scored

    Full Stack System Design (Backend + Frontend )-40%AI Integration & Safety Thinking -25%Ownership, Communication, & Pragmatism - 25%AI-Assisted Development Perspective - 10%

    What to Submit

    Please submit the following:

    1. Video Recording (≤ 10 minutes):
    • One continuous recording addressing both Part 1 (Personal Context) and Part 2 (Challenge Scenario)
    • Ensure audio and video quality are clear
    • You may use screen sharing to show diagrams or notes
    1. Written Artifacts:
    • Any notes, diagrams, or sketches you prepared for your video
    • Can be hand-drawn, digital, or text-based
    1. AI Usage Declaration:

    At the end of your video or in a brief note, please include:

    ☐ I did not use any AI tools to prepare my answer

    ☐ I used AI tools (briefly describe how --- for example, "I used AI to help organize my talking points, then rewrote everything in my own words.")

    Video Recording

    Video · 10 min maxRequired

    Format: .mp4, .mov, .webm

    Part 1: Personal Scenario (≈2 minutes)

    TalentReach wants to understand your broader career goals. If this role isn't the perfect fit, we'd love to help point you toward other opportunities.

    Share some or all of these:

    • Why this role? What excites you about it?
    • If not this role, what other roles interest you?
    • Company environment: startup, growth-stage, or enterprise?
    • Culture fit: What matters most? (innovation, stability, mentorship, diversity)
    • Career trajectory: IC or leadership path?
    • Contract, permanent, or both?
    • Salary range or hourly rate
    • Remote, hybrid, or in-office preference
    • Travel requirements acceptable?
    • Tech stack preferences
    • Product development vs. infrastructure
    • Greenfield vs. maintenance
    • Preferred methodologies (Agile, Scrum, etc.)
    • Non-negotiables
    • Visa sponsorship needs

    Part 2: Challenge Scenario (≈8 minutes)

    Walk us through:

    A. System & Data Flow (backend + frontend)

    • How would you structure the flow from "new message arrives" → "AI proposes actions" → "user approves and executes"?
    • Key backend components (Django services, APIs, data stores) and key React pieces (views, state management)
    • How would you keep this first version simple but scalable?

    B. Safety, Reliability, and Human-in-the-Loop

    • How would you ensure the AI assistant doesn't do anything unsafe or surprising?
    • Where do humans approve or override AI suggestions?
    • What logging or audit trail would you put in place?

    C. Using AI-Assisted Development Tools

    • How would you use tools like Cursor, Windsurf, or similar to speed up building this feature without treating them as a black box?
    • Give one concrete example and explain when you trust AI output vs. when you dig deeper yourself.

    Sign in to upload files

    Written Artifacts

    DocumentOptional

    Format: .pdf, .doc, .docx, .rtf, .txt, .md

    Any notes, diagrams, or sketches prepared for your video. Can be hand-drawn (scanned/photographed), digital, or text-based.

    Sign in to upload files