Upsonic

<div align="center"> <img src="https://github.com/user-attachments/assets/fbe7219f-55bc-4748-ac4a-dd2fb2b8d9e5" width="600" /> # Upsonic **Production-Ready AI Agent Framework with Safety First** [](https://badge.fury.io/py/upsonic) [](LICENCE) [](https://pypi.org/project/upsonic/) [](https://github.com/Upsonic/Upsonic) [](https://github.com/Upsonic/Upsonic/issues) [](https://docs.upsonic.ai) [](https://discord.gg/pmYDMSQHqY) [Documentation](https://docs.upsonic.ai) • [Quickstart](https://docs.upsonic.ai/get-started/quickstart) • [Examples](https://docs.upsonic.ai/examples) • [Discord](https://discord.gg/pmYDMSQHqY) </div> --- ## Overview Upsonic is an open-source AI agent framework for building production-ready agents. It supports multiple AI providers (OpenAI, Anthropic, Azure, Bedrock) and includes built-in safety policies, OCR, memory, multi-agent coordination, and MCP tool integration. ## What Can You Build? - **Document Analysis**: Extract and process text from images and PDFs - **Customer Service Automation**: Agents with memory and session context - **Financial Analysis**: Agents that analyze data, generate reports, and provide insights - **Compliance Monitoring**: Enforce safety policies across all agent interactions - **Research & Data Gathering**: Automate research workflows with multi-agent collaboration - **Multi-Agent Workflows**: Orchestrate tasks across specialized agent teams ## Quick Start ### Installation ```bash uv pip install upsonic # pip install upsonic ``` ### Basic Agent ```python from upsonic import Agent, Task agent = Agent(model="anthropic/claude-sonnet-4-5", name="Stock Analyst Agent") task = Task(description="Analyze the current market trends") agent.print_do(task) ``` ### Agent with Tools ```python from upsonic import Agent, Task from upsonic.tools.common_tools import YFinanceTools agent = Agent(model="anthropic/claude-sonnet-4-5", name="Stock Analyst Agent") task = Task( description="Give me a summary about tesla stock with tesla car models", tools=[YFinanceTools()] ) agent.print_do(task) ``` ### Agent with Memory ```python from upsonic import Agent, Task from upsonic.storage import Memory, InMemoryStorage memory = Memory( storage=InMemoryStorage(), session_id="session_001", full_session_memory=True ) agent = Agent(model="anthropic/claude-sonnet-4-5", memory=memory) task1 = Task(description="My name is John") agent.print_do(task1) task2 = Task(description="What is my name?") agent.print_do(task2) # Agent remembers: "Your name is John" ``` **Ready for more?** Check out the [Quickstart Guide](https://docs.upsonic.ai/get-started/quickstart) for additional examples including Knowledge Base and Team workflows. ## Key Features - **Autonomous Agent**: An agent that can read, write, and execute code inside a sandboxed workspace, no tool setup required - **Safety Engine**: Policy-based content filtering applied to user inputs, agent outputs, and tool interactions - **OCR Support**: Unified interface for multiple OCR engines with PDF and image support - **Memory Management**: Session memory and long-term storage with multiple backend options - **Multi-Agent Teams**: Sequential and parallel agent coordination - **Tool Integration**: MCP tools, custom tools, and human-in-the-loop workflows - **Production Ready**: Monitoring, metrics, and enterprise deployment support ## Core Capabilities ### Autonomous Agent `AutonomousAgent` extends `Agent` with built-in filesystem and shell tools, automatic session memory, and workspace sandboxing. Useful for coding assistants, DevOps automation, and any task that needs direct file or terminal access. ```python from upsonic import AutonomousAgent, Task agent = AutonomousAgent( model="anthropic/claude-sonnet-4-5", workspace="/path/to/project" ) task = Task("Read the main.py file and add error handling to every function") agent.print_do(task) ``` All file and shell operations are restricted to `workspace`. Path traversal and dangerous commands are blocked. --- ### Safety Engine The Safety Engine applies policies at three points: user inputs, agent outputs, and tool interactions. Policies can block, anonymize, replace, or raise exceptions on matched content. ```python from upsonic import Agent, Task from upsonic.safety_engine.policies.pii_policies import PIIAnonymizePolicy agent = Agent( model="anthropic/claude-sonnet-4-5", user_policy=PIIAnonymizePolicy, # anonymizes PII before sending to the LLM ) task = Task( description="My email is john.doe@example.com and phone is 555-1234. What are my email and phone?" ) # PII is anonymized before reaching the LLM, then de-anonymized in the response result = agent.do(task) print(result) # "Your email is john.doe@example.com and phone is 555-1234" ``` Pre-built policies cover PII, adult content, profanity, financial data, and more. Custom policies are also supported. Learn more: [Safety Engine Documentation](https://docs.upsonic.ai/concepts/safety-engine/overview) --- ### OCR and Document Processing Upsonic provides a unified OCR interface with a layered pipeline: Layer 0 handles document preparation (PDF to image conversion, preprocessing), Layer 1 runs the OCR engine. ```bash uv pip install "upsonic[ocr]" ``` ```python from upsonic.ocr import OCR from upsonic.ocr.layer_1.engines import EasyOCREngine engine = EasyOCREngine(languages=["en"]) ocr = OCR(layer_1_ocr_engine=engine) text = ocr.get_text("invoice.pdf") print(text) ``` Supported engines: EasyOCR, RapidOCR, Tesseract, PaddleOCR, DeepSeek OCR, DeepSeek via Ollama. Learn more: [OCR Documentation](https://docs.upsonic.ai/concepts/ocr/overview) ## Upsonic AgentOS AgentOS is an optional deployment platform for running agents in production. It provides a Kubernetes-based runtime, metrics dashboard, and self-hosted deployment. - **Kubernetes-based FastAPI Runtime**: Deploy agents as isolated, scalable microservices - **Metrics Dashboard**: Track LLM costs, token usage, and performance per transaction - **Self-Hosted**: Full control over your data and infrastructure - **One-Click Deployment**: Automated deployment pipelines <img width="3024" height="1590" alt="AgentOS Dashboard" src="https://github.com/user-attachments/assets/42fceaca-2dec-4496-ab67-4b9067caca42" /> ## Documentation and Resources - **[Documentation](https://docs.upsonic.ai)** - Complete guides and API reference - **[Quickstart Guide](https://docs.upsonic.ai/get-started/quickstart)** - Get started in 5 minutes - **[Examples](https://docs.upsonic.ai/examples)** - Real-world examples and use cases - **[API Reference](https://docs.upsonic.ai/reference)** - Detailed API documentation ## Community and Support > **💬 [Join our Discord community!](https://discord.gg/pmYDMSQHqY)** — Ask questions, share what you're building, get help from the team, and connect with other developers using Upsonic. - **[Discord](https://discord.gg/pmYDMSQHqY)** - Chat with the community and get real-time support - **[Issue Tracker](https://github.com/Upsonic/Upsonic/issues)** - Report bugs and request features - **[Changelog](https://docs.upsonic.ai/changelog)** - See what's new in each release ## License Upsonic is released under the MIT License. See [LICENCE](LICENCE) for details. ## Contributing We welcome contributions from the community! Please read our contributing guidelines and code of conduct before submitting pull requests. --- **Learn more at [upsonic.ai](https://upsonic.ai)**