A five-level maturity model for AI-human collaboration with anticipatory empathy
Project description
Empathy
AI that predicts problems before they happen.
Most AI tools are reactive - they wait for you to ask, then respond. Empathy is anticipatory - it predicts what you'll need and warns you before problems happen.
pip install empathy
What It Does
- 🔮 Predicts issues 30-90 days ahead - Security vulnerabilities, performance bottlenecks, compliance gaps
- 🧠 Learns patterns across domains - Healthcare handoff protocols → deployment safety checks
- 🔌 Works with any LLM - Claude, GPT-4, Gemini, local models via Ollama
- 🏥 Enterprise-ready - PII scrubbing, audit logging, HIPAA/GDPR compliant
- 📦 2,000+ downloads on PyPI, 2,040+ tests passing
Quick Example
from empathy_os import EmpathyOS
os = EmpathyOS()
# Analyze code for current AND future issues
result = await os.collaborate(
"Review this deployment pipeline for problems",
context={"code": pipeline_code, "team_size": 10}
)
# Get predictions, not just analysis
print(result.current_issues) # What's wrong now
print(result.predicted_issues) # What will break in 30-90 days
print(result.prevention_steps) # How to prevent it
The 5 Levels of AI Empathy
| Level | Name | Behavior | Example |
|---|---|---|---|
| 1 | Reactive | Responds when asked | "Here's the data you requested" |
| 2 | Guided | Asks clarifying questions | "What format do you need?" |
| 3 | Proactive | Notices patterns | "I pre-fetched what you usually need" |
| 4 | Anticipatory | Predicts future needs | "This query will timeout at 10k users" |
| 5 | Transformative | Builds preventing structures | "Here's a framework for all future cases" |
Empathy operates at Level 4 - predicting problems before they manifest.
Why Empathy?
| Empathy | SonarQube | GitHub Copilot | |
|---|---|---|---|
| Predicts future issues | ✅ 30-90 days ahead | ❌ | ❌ |
| Cross-domain learning | ✅ Healthcare → Software | ❌ | ❌ |
| Source available | ✅ Fair Source 0.9 | ❌ | ❌ |
| Free for small teams | ✅ ≤5 employees | ❌ | ❌ |
| Local/air-gapped | ✅ Ollama support | ❌ | ❌ |
Get Involved
⭐ Star this repo if you find it useful
💬 Join Discussions - Questions, ideas, show what you built
📖 Read the Book - Deep dive into the philosophy and implementation
📚 Full Documentation - API reference, examples, guides
Install Options
# Basic
pip install empathy
# With all features (recommended)
pip install empathy[full]
# Development
git clone https://github.com/Smart-AI-Memory/empathy.git
cd empathy && pip install -e .[dev]
What's Included
- 30+ Wizards - Security, performance, testing, accessibility, compliance
- Healthcare Suite - SBAR, SOAP notes, clinical protocols (HIPAA compliant)
- IDE Plugins - VS Code and JetBrains extensions (examples/)
- Enterprise Security - PII scrubbing, secrets detection, audit logging
License
Fair Source License 0.9 - Free for students, educators, and teams ≤5 employees. Commercial license ($99/dev/year) for larger organizations. Details →
Built by Smart AI Memory · Documentation · Examples · Issues
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file empathy_framework-2.1.0.tar.gz.
File metadata
- Download URL: empathy_framework-2.1.0.tar.gz
- Upload date:
- Size: 2.0 MB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.11
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
cbcbea5db2f811503b160eb16cb7bcd62c23afa4c3fbc38c03f4cd18aa830914
|
|
| MD5 |
07deeff1e095c27800da6a82ca90c9a2
|
|
| BLAKE2b-256 |
3782df146b48e8369f11a9a96ae686a32506326f8ea37bfcfd3614bc72e098c2
|
File details
Details for the file empathy_framework-2.1.0-py3-none-any.whl.
File metadata
- Download URL: empathy_framework-2.1.0-py3-none-any.whl
- Upload date:
- Size: 247.2 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.11
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
9703f35446ebe34ce183c130bc94d7be155bb163d6548f459d001d0edc9ee467
|
|
| MD5 |
0ce7e2afb0b5ac707227a03021d6d662
|
|
| BLAKE2b-256 |
a6d0ded7ee3eeb092490356113e13c58462e01fad89ee7465ccb9c0acd099cd3
|