# Project export: Meredith This document was generated by HackStack to give an AI agent context about a hackathon project. Sections are labeled with their provenance; content marked as truncated was cut to keep this document small. ## Project metadata - Hackathon: TreeHacks 2025 - Tagline: A fully autonomous agent powered by edge AI technology on your Apple Watch, dynamically planning complex workflows across email, calendar, files, and Perplexity search - all while keeping data private - Devpost: https://devpost.com/software/meredith - GitHub: https://github.com/xzrderek/treehacks25/ - Video: https://www.youtube.com/embed/d-OK8ggtVjU?enablejsapi=1&hl=en_US&rel=0&start=&version=3&wmode=transparent - Team: 2 GitHub contributor(s) — Derek Xu (15 commits), Abhinav Kommula (3 commits) ## Devpost submission (written by the team) ### Inspiration Medical residents juggle the impossible: 80+ hour weeks, nonstop learning, and high-stakes decisions, all while battling fatigue. Our team was inspired by family and friends in residency, sharing stories of 28-hour shifts, only to attend rounds, study for exams, and prep for procedures with barely a moment to breathe. AI assistants like ChatGPT and Claude offer powerful tools, but sending sensitive medical data to the cloud is a non-starter in healthcare. Residents need real support without compromising HIPAA compliance or patient privacy. That’s why we built Meredith: an AI-powered agent that connects to your Apple Watch, processing everything locally so clinical notes, schedules, and patient data never leave your computer. No cloud uploads. No privacy concerns. Just an intelligent, hands-free partner to help lighten the load of residency, right from your wrist. With Meredith, you’re not just surviving residency. You’re mastering it. ### What it does Meredith serves as an intelligent companion for medical residents, accessible through voice commands on their Apple Watch. It streamlines daily tasks by managing schedules, facilitating communication, retrieving patient information, and researching internet sources--all while maintaining strict medical privacy standards. The core functionality includes smart scheduling, automated documentation in standard medical formats (SBAR, admission notes, procedure notes, discharge summaries), instant access to up-to-date medical research, and intelligent task management that adapts to the dynamic nature of hospital work. ### How we built it Unlike traditional AI assistants that rely on cloud processing, we developed Meredith using an Edge AI architecture so that language model inference occurs locally on our computers. This approach ensures complete patient data privacy and HIPAA compliance--no sensitive information ever leaves the device. The system is powered by an existing open-source fine-tuned model, TinyAgent-7B, that operates directly on the device via a llama.cpp HTTP server, integrating seamlessly with essential APIs: Perplexity Sonar for real-time medical research Native iOS APIs (Contacts, SMS, Email, Calendar, Maps, File Management) To develop the actual agent, our system builds a DAG (Directed Acyclic Graph) for task planning and execution. When given instructions, Meredith determines the optimal task ordering to build the function calling DAG where each node represents an API (like sending an email, or creating a calendar event) and directed edges represent dependencies between tasks (for example, needing to fetch email addresses before sending a meeting invite). This DAG-based planning ensures tasks are executed in the correct sequence while maximizing parallel execution where possible. Our agentic architecture also enables the flexible integration of new tools and APIs through in-context learning and prompt engineering. On the full stack end, we built our web app using Next.js, Flask, and Python. Our Apple Watch app was built in Xcode using Swift. ### Challenges we ran into Adding custom tooling for the TinyAgent model. The model was originally fine-tuned to pick between 16 different tools, which made adding additional functionality impossible without retraining the last layer of the model. We solved this via prompt engineering and examples so that the LLM could leverage its ICL capabilities to choose the most relevant tools. Guiding the agent to execute more complex workflows was also difficult, as the TinyAgent model often would get stuck if a single API call resulted in an error. ### Accomplishments we're proud of Successfully implementing a task-planning agentic system powered by on edge AI! Building a full-stack application using Typescript, Next.js, and Flask (we're all backend engineers) Spinning up the Apple Watch task queue app, and enabling it to convert voice-to-text ### What's next Meredith's potential extends far beyond its current capabilities. Paramedics operate in high-stress environments where every second counts and their hands are often occupied with life-saving procedures. A voice-first application would allow them to maintain situational awareness while documenting critical information and coordinating care. Imagine a paramedic arriving at a trauma scene: they could verbally log their arrival time, document initial patient vitals, and alert the receiving hospital about incoming trauma cases, all while physically attending to the patient. Meredith could transcribe their verbal notes into structured medical documentation, set up automated alerts to the emergency department, and maintain a real-time log of interventions and medications administered. Additionally, many of the potential use cases of Meredith operate in high-stakes medical scenarios, where errors could impact patient care decisions or critical diagnoses. We want Meredith to be as useful as possible, but building robust safety guardrails will be a significant blocker in its adoption by medical residents. ## README (from the GitHub repository) # treehacks25 ## Detected evidence (automated analysis) Indexed codebase: 128 recognized source files, 724 KB. - CSS (language) — detected in the code - FastAPI (technology) — detected in the code - Flask (technology) — detected in the code - Hugging Face (technology) — detected in the code - JavaScript (language) — detected in the code - LangChain (technology) — detected in the code - Next.js (technology) — detected in the code - OpenAI (technology) — detected in the code - Python (language) — detected in the code - PyTorch (technology) — detected in the code - React (technology) — detected in the code - Swift (language) — detected in the code - Tailwind CSS (technology) — detected in the code - TypeScript (language) — detected in the code ## Codebase structure (from repository index) ### Files (120 of 163) ``` .gitignore app/globals.css app/layout.tsx app/page.tsx app/tasks/[taskId]/page.tsx backend/__init__.py backend/dag.py backend/index.py backend/models.py backend/pyproject.toml backend/README.md backend/routes.py backend/run.py backend/services.py backend/task_queue.py backend/uv.lock components.json components/ui/accordion.tsx components/ui/alert-dialog.tsx components/ui/alert.tsx components/ui/aspect-ratio.tsx components/ui/avatar.tsx components/ui/badge.tsx components/ui/breadcrumb.tsx components/ui/button.tsx components/ui/calendar.tsx components/ui/card.tsx components/ui/carousel.tsx components/ui/chart.tsx components/ui/checkbox.tsx components/ui/collapsible.tsx components/ui/command.tsx components/ui/context-menu.tsx components/ui/dialog.tsx components/ui/drawer.tsx components/ui/dropdown-menu.tsx components/ui/form.tsx components/ui/hover-card.tsx components/ui/input-otp.tsx components/ui/input.tsx components/ui/label.tsx components/ui/menubar.tsx components/ui/navigation-menu.tsx components/ui/pagination.tsx components/ui/popover.tsx components/ui/progress.tsx components/ui/radio-group.tsx components/ui/resizable.tsx components/ui/scroll-area.tsx components/ui/select.tsx components/ui/separator.tsx components/ui/sheet.tsx components/ui/sidebar.tsx components/ui/skeleton.tsx components/ui/slider.tsx components/ui/sonner.tsx components/ui/switch.tsx components/ui/table.tsx components/ui/tabs.tsx components/ui/textarea.tsx components/ui/toast.tsx components/ui/toaster.tsx components/ui/toggle-group.tsx components/ui/toggle.tsx components/ui/tooltip.tsx config.json hooks/use-mobile.tsx hooks/use-toast.ts lib/utils.ts next.config.js package.json pineapple/pineapple WatchKit App/Assets.xcassets/AccentColor.colorset/Contents.json pineapple/pineapple WatchKit App/Assets.xcassets/AppIcon.appiconset/Contents.json pineapple/pineapple WatchKit App/Assets.xcassets/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Circular.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Extra Large.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Graphic Bezel.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Graphic Circular.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Graphic Corner.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Graphic Extra Large.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Graphic Large Rectangular.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Modular.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Complication.complicationset/Utilitarian.imageset/Contents.json pineapple/pineapple WatchKit Extension/Assets.xcassets/Contents.json pineapple/pineapple WatchKit Extension/ComplicationController.swift pineapple/pineapple WatchKit Extension/Info.plist pineapple/pineapple WatchKit Extension/NotificationController.swift pineapple/pineapple WatchKit Extension/NotificationView.swift pineapple/pineapple WatchKit Extension/pineappleApp.swift pineapple/pineapple WatchKit Extension/Preview Content/Preview Assets.xcassets/Contents.json pineapple/pineapple WatchKit Extension/PushNotificationPayload.apns pineapple/pineapple WatchKit Extension/View/ContentView.swift pineapple/pineapple WatchKit Extension/View/DetailView.swift pineapple/pineapple WatchKit Extension/View/Model/Task.swift pineapple/pineapple.xcodeproj/project.pbxproj pineapple/pineapple.xcodeproj/project.xcworkspace/contents.xcworkspacedata pineapple/pineapple.xcodeproj/project.xcworkspace/xcshareddata/IDEWorkspaceChecks.plist pineapple/pineapple.xcodeproj/project.xcworkspace/xcuserdata/derekxu.xcuserdatad/UserInterfaceState.xcuserstate pineapple/pineapple.xcodeproj/project.xcworkspace/xcuserdata/rohanviswanathan.xcuserdatad/UserInterfaceState.xcuserstate pineapple/pineapple.xcodeproj/xcuserdata/derekxu.xcuserdatad/xcschemes/xcschememanagement.plist pineapple/pineapple.xcodeproj/xcuserdata/rohanviswanathan.xcuserdatad/xcschemes/xcschememanagement.plist pineapple/pineappleTests/pineappleTests.swift pineapple/pineappleUITests/pineappleUITests.swift pineapple/pineappleUITests/pineappleUITestsLaunchTests.swift postcss.config.js README.md requirements.txt tailwind.config.js tinyagent/.gitignore tinyagent/pyproject.toml tinyagent/README.md tinyagent/requirements.txt tinyagent/run_tiny_agent_server.py tinyagent/setup.py tinyagent/src/agents/agent.py tinyagent/src/agents/structured_chat_agent.py tinyagent/src/agents/tools.py tinyagent/src/callbacks/callbacks.py tinyagent/src/chains/chain.py [43 more files omitted for size] ``` ### Dependencies - package.json: @hookform/resolvers@^4.0.0, @radix-ui/react-accordion@^1.2.3, @radix-ui/react-alert-dialog@^1.1.6, @radix-ui/react-aspect-ratio@^1.1.2, @radix-ui/react-avatar@^1.1.3, @radix-ui/react-checkbox@^1.1.4, @radix-ui/react-collapsible@^1.1.3, @radix-ui/react-context-menu@^2.2.6, @radix-ui/react-dialog@^1.1.6, @radix-ui/react-dropdown-menu@^2.1.6, @radix-ui/react-hover-card@^1.1.6, @radix-ui/react-label@^2.1.2, @radix-ui/react-menubar@^1.1.6, @radix-ui/react-navigation-menu@^1.2.5, @radix-ui/react-popover@^1.1.6, @radix-ui/react-progress@^1.1.2, @radix-ui/react-radio-group@^1.2.3, @radix-ui/react-scroll-area@^1.2.3, @radix-ui/react-select@^2.1.6, @radix-ui/react-separator@^1.1.2, @radix-ui/react-slider@^1.2.3, @radix-ui/react-slot@^1.1.2, @radix-ui/react-switch@^1.1.3, @radix-ui/react-tabs@^1.1.3, @radix-ui/react-toast@^1.2.6, @radix-ui/react-toggle@^1.1.2, @radix-ui/react-toggle-group@^1.1.2, @radix-ui/react-tooltip@^1.1.8, @types/node@20.1.4, @types/react@18.2.6, @types/react-dom@18.2.4, autoprefixer@10.4.14, class-variance-authority@^0.7.1, clsx@^2.1.1, cmdk@1.0.0, concurrently@^8.0.1, date-fns@^3.0.0, embla-carousel-react@^8.5.2, eslint@8.40.0, eslint-config-next@13.4.2, input-otp@^1.4.2, lucide-react@^0.475.0, next@^15.1.7, next-themes@^0.4.4, postcss@^8.5.2, react@18.2.0, react-day-picker@8.10.1, react-dom@18.2.0, react-hook-form@^7.54.2, react-markdown@^9.0.3, react-resizable-panels@^2.1.7, recharts@^2.15.1, sonner@^1.7.4, tailwind-merge@^3.0.1, tailwindcss@3.3.2, tailwindcss-animate@^1.0.7, typescript@5.0.4, vaul@^1.1.2, zod@^3.24.2 - requirements.txt: Flask@==3.0.3, Flask-CORS@==4.0.0, Flask-Migrate@==4.0.4, Flask-SQLAlchemy@==3.0.5 - tinyagent/requirements.txt: bs4@==0.0.1, fastapi@==0.110.1, httpx@==0.27.0, langchain@==0.1.0, langchain-community@==0.0.9, langchain-core@==0.1.7, langchain-openai@==0.0.2, numexpr@==2.8.7, openai@==1.6.1, protobuf@==5.26.1, PyMuPDF@==1.23.26, python-dateutil@==2.8.2, python-multipart@==0.0.9, sentencepiece@==0.2.0, tiktoken@==0.5.2, torch@==2.2.2, transformers@==4.38.2, uvicorn@==0.29.0 ### Recent commits (newest first) - fixing config file, using local llama cpp - update readme - final changes - accidentally deleted tinyagent - finished backend now, working - lots of backend progress - renaming folders - Added AppleWatch backend and fixed components - Modified frontend to visualize task progress - Fixed frontend and added backend changes - draft of UI - delete test_tinyagent.py - Redo tinyagent setup - queue implementation - turned tinyagent into a package - adding tinyagent to repo - base flask nextjs webapp running - first commit ## Key source files (fetched from GitHub, selected and truncated for size) ### requirements.txt ``` Flask==3.0.3 Flask-SQLAlchemy==3.0.5 Flask-Migrate==4.0.4 Flask-CORS==4.0.0 ``` ### package.json ``` { "name": "next-flask", "version": "0.1.0", "private": true, "scripts": { "flask-dev": "FLASK_DEBUG=1 pip3 install -r requirements.txt && python3 -m flask --app backend/index run -p 8000 --host=0.0.0.0", "next-dev": "next dev", "dev": "concurrently \"pnpm run next-dev\" \"pnpm run flask-dev\"", "build": "next build", "start": "next start", "lint": "next lint" }, "dependencies": { "@hookform/resolvers": "^4.0.0", "@radix-ui/react-accordion": "^1.2.3", "@radix-ui/react-alert-dialog": "^1.1.6", "@radix-ui/react-aspect-ratio": "^1.1.2", "@radix-ui/react-avatar": "^1.1.3", "@radix-ui/react-checkbox": "^1.1.4", "@radix-ui/react-collapsible": "^1.1.3", "@radix-ui/react-context-menu": "^2.2.6", "@radix-ui/react-dialog": "^1.1.6", "@radix-ui/react-dropdown-menu": "^2.1.6", "@radix-ui/react-hover-card": "^1.1.6", "@radix-ui/react-label": "^2.1.2", "@radix-ui/react-menubar": "^1.1.6", "@radix-ui/react-navigation-menu": "^1.2.5", "@radix-ui/react-popover": "^1.1.6", "@radix-ui/react-progress": "^1.1.2", "@radix-ui/react-radio-group": "^1.2.3", "@radix-ui/react-scroll-area": "^1.2.3", "@radix-ui/react-select": "^2.1.6", "@radix-ui/react-separator": "^1.1.2", "@radix-ui/react-slider": "^1.2.3", "@radix-ui/react-slot": "^1.1.2", "@radix-ui/react-switch": "^1.1.3", "@radix-ui/react-tabs": "^1.1.3", "@radix-ui/react-toast": "^1.2.6", "@radix-ui/react-toggle": "^1.1.2", "@radix-ui/react-toggle-group": "^1.1.2", "@radix-ui/react-tooltip": "^1.1.8", "@types/node": "20.1.4", "@types/react": "18.2.6", "@types/react-dom": "18.2.4", "autoprefixer": "10.4.14", "class-variance-authority": "^0.7.1", "clsx": "^2.1.1", "cmdk": "1.0.0", "concurrently": "^8.0.1", "date-fns": "^3.0.0", "embla-carousel-react": "^8.5.2", "eslint": "8.40.0", "eslint-config-next": "13.4.2", "input-otp": "^1.4.2", "lucide-react": "^0.475.0", "next": "^15.1.7", "next-themes": "^0.4.4", "postcss": "^8.5.2", "react": "18.2.0", "react-day-picker": "8.10.1", "react-dom": "18.2.0", "react-hook-form": "^7.54.2", "react-markdown": "^9.0.3", "react-resizable-panels": "^2.1.7", "recharts": "^2.15.1", "sonner": "^1.7.4", "tailwind-merge": "^3.0.1", "tailwindcss": "3.3.2", "tailwindcss-animate": "^1.0.7", "typescript": "5.0.4", "vaul": "^1.1.2", "zod": "^3.24.2" } } ``` ### backend/pyproject.toml ``` [project] name = "backend" version = "0.1.0" description = "Add your description here" readme = "README.md" requires-python = ">=3.11" dependencies = [] [build-system] requires = ["hatchling"] build-backend = "hatchling.build" ``` ### tinyagent/pyproject.toml ``` [project] name = "tinyagent" version = "0.1.0" description = "Add your description here" readme = "README.md" requires-python = ">=3.12" dependencies = [] [build-system] requires = ["hatchling"] build-backend = "hatchling.build" ``` ### tinyagent/requirements.txt ``` PyMuPDF==1.23.26 bs4==0.0.1 fastapi==0.110.1 langchain-community==0.0.9 langchain-core==0.1.7 langchain-openai==0.0.2 langchain==0.1.0 numexpr==2.8.7 openai==1.6.1 protobuf==5.26.1 python-dateutil==2.8.2 sentencepiece==0.2.0 tiktoken==0.5.2 torch==2.2.2 transformers==4.38.2 uvicorn==0.29.0 python-multipart==0.0.9 httpx==0.27.0 ``` ### app/layout.tsx ```typescript import './globals.css' import { Inter } from 'next/font/google' const inter = Inter({ subsets: ['latin'] }) export const metadata = { title: 'Task Manager', description: 'A simple task execution dashboard', } export default function RootLayout({ children, }: { children: React.ReactNode }) { return (
{children}
) } ``` ### backend/index.py ```python from flask import Flask from .routes import tinyagent_bp from flask_sqlalchemy import SQLAlchemy from flask_migrate import Migrate import enum from flask_cors import CORS from . import task_queue # Add this import # db = SQLAlchemy() migrate = Migrate() app = Flask(__name__) CORS(app) app.register_blueprint(tinyagent_bp) # db.init_app(app) # migrate.init_app(app, db) # The task queue worker thread will start automatically when imported @app.route("/") def home(): return "Flask server is running!", 200 @app.route("/api/python") def hello_world(): return "

Hello, World!

" if __name__ == '__main__': app.run(host='0.0.0.0', port=5000, debug=True) ``` ### app/page.tsx ```typescript "use client"; import { useState, useEffect } from "react"; import { useRouter } from 'next/navigation'; interface Task { task_id: string; task_description: string; status: string; date: string; result?: string; } export default function Home() { const [tasks, setTasks] = useState([]); const [newTask, setNewTask] = useState(""); const [mounted, setMounted] = useState(false); const router = useRouter(); const fetchTasks = async () => { try { const response = await fetch(`${process.env.NEXT_PUBLIC_API_URL}/tasks`); const data = await response.json(); const updatedTasks = Object.values(data.tasks).map((task: any) => ({ task_id: task.task_id, task_description: task.task_description, status: task.status.charAt(0).toUpperCase() + task.status.slice(1), date: task.date, result: task.result })); setTasks(updatedTasks); } catch (error) { console.error("Error fetching tasks:", error); } }; useEffect(() => { setMounted(true); fetchTasks(); const interval = setInterval(fetchTasks, 2000); return () => clearInterval(interval); }, []); if (!mounted) { return null; } const handleAddTask = async () => { if (!newTask.trim()) return; try { const response = await fetch(`${process.env.NEXT_PUBLIC_API_URL}/tasks/submit`, { method: "POST", headers: { "Content-Type": "application/json" }, body: JSON.stringify({ query: newTask }), }); if (!response.ok) throw new Error("Failed to add task"); const newTaskData = await response.json(); setTasks([...tasks, { task_id: newTaskData.task_id, task_description: newTask, status: "Pending", date: "2024-02-14 12:00:00", // Fixed date for testing }]); setNewTask(""); } catch (error) { console.error("Error adding task:", error); } }; const handleDeleteTask = async (taskId: string) => { try { const response = await fetch(`${process.env.NEXT_PUBLIC_API_URL}/tasks/${taskId}`, { method: 'DELETE', }); if (!response.ok) throw new Error('Failed to delete task'); setTasks(tasks.filter(task => task.task_id !== taskId)); } catch (error) { console.error('Error deleting task:', error); } }; const getStatusColor = (status: string) => { switch (status) { case "Completed": return "bg-green-200 text-green-700"; case "Processing": return "bg-yellow-200 text-yellow-700"; case "Failed": return "bg-red-200 text-red-700"; default: return "bg-gray-200 text-gray-700"; } }; return (
{/* Header with Task Input */}

Hello, I'm Meredith

setNewTask(e.target.value)} onKeyDown={(e) => { if (e.key === 'Enter') { handleAddTask(); } }} className="flex-1 px-4 py-3 border border-gray-300 rounded-lg focus:outline-none focus:ring-2 focus:ring-blue-500" />
{/* Task Queue */}

Task Queue

{tasks.map((task) => (
router.push(`/tasks/${task.task_id}`)} className="bg-white rounded-lg p-4 shadow-sm hover:shadow-md transition cursor-pointer border border-gray-100" >
{task.status}

{task.task_description}

{task.date}

))} {tasks.length === 0 && (

Agent Scratchpad

{log.agent_scratchpad.split('\n').map((line, idx) => (
{line}
))}

Planner Response

{log.planner_response.split('\n').map((step, idx) => (
{idx + 1}
{step}
))}

Task Log

{log.task_log.split('\n').map((entry, idx) => (
{entry}
))}
); }; const StatusBadge = ({ status }: { status: string }) => { const getStatusColor = () => { switch (status.toLowerCase()) { case 'completed': return 'bg-green-100 text-green-800'; case 'failed': return 'bg-red-100 text-red-800'; case 'running': return 'bg-blue-100 text-blue-800'; default: return 'bg-gray-100 text-gray-800'; } }; const getStatusIcon = () => { switch (status.toLowerCase()) { case 'completed': return ; case 'failed': return ; case 'running': return ; default: return null; } }; return ( {getStatusIcon()} {status} ); }; export default function TaskPage() { const params = useParams(); const taskId = params.taskId as string; const [task, setTask] = useState(null); const [loading, setLoading] = useState(true); const [error, setError] = useState(null); useEffect(() => { const fetchTaskDetails = async () => { if (!taskId) return; try { const response = await fetch(`${process.env.NEXT_PUBLIC_API_URL}/tasks/${taskId}`); if (!response.ok) { throw new Error('Failed to fetch task details'); } const data = await response.json(); setTask(data); } catch (err) { setError(err instanceof Error ? err.message : 'Failed to fetch task details'); } finally { setLoading(false); } }; fetchTaskDetails(); }, [taskId]); if (loading) { return
Loading...
; } if (error) { return
Error: {error}
; } if (!task) { return
Task not found
; } return (

Task Details

Description

{task.task_description}

Timing

Started: {new Date(task.started_at).toLocaleString()}
Completed: {new Date(task.completed_at).toLocaleString()}

Metadata

Agent Version: {task.metadata.agent_version}
Priority: {task.metadata.priority}
Retries: {task.metadata.retries}
{task.result && (

Result

{task.result}
)} {task.error_message && (