LangWatch: The Open LLM Ops Platform for Confident AI Development

Want to build robust and reliable AI applications? Discover LangWatch, the open LLM Ops platform designed to help you track, visualize, and analyze your LLM interactions with ease. Learn how to leverage its powerful features for enhanced usability and performance.

Why LangWatch for Your LLM Applications?

LangWatch isn't just another tool; it's a comprehensive suite meticulously designed to elevate how you interact with and understand your language models. By centralizing key functionalities like debugging, evaluation, and real-time monitoring, LangWatch empowers both technical and non-technical teams to work collaboratively towards creating top-tier LLM applications. Below are some benefits:

• Enhanced Collaboration: Streamline teamwork with a unified platform.
• Improved Performance: Fine-tune your LLM applications for optimal results.
• Deeper Insights: Gain a thorough understanding of user engagement.

Key Features of the LangWatch LLMOps Platform

LangWatch offers a range of features designed to provide insights and improve LLM application quality. From real-time telemetry to automated DSPy prompt optimization, explore the functionalities that make LangWatch a must-have tool.

• Real-time Telemetry: Detailed interaction tracings for LLM cost and latency optimization.
• Detailed Debugging: Comprehensive capture of LLM call chains with metadata, grouped for easy troubleshooting.
• Measurable LLM Quality: Use LangEvals evaluators to quantitatively measure pipeline output quality.

Visualize and Optimize with the DSPy Visualizer

Optimize prompts and pipelines effortlessly with LangWatch's DSPy Visualizer. Track progress, compare runs, and iterate towards perfection.

• Easy Inspection: Easily inspect and track your DSPy experiments.
• Historical Data: Maintain a history of your DSPy runs for comparison.
• Progress Tracking: Follow the progress of your experiments in real-time.

User Analytics & Guardrails: Understand and Protect Your Users

Dive into behavior patterns and message insights to understand user engagement, and implement guardrails to secure LLM outputs.

• User Analytics: Metrics on engagement and user interactions for product improvement.
• Guardrails: Detect PII leaks and toxic language with built-in and custom guardrails.
• Customizable Alerts: Trigger alerts based on semantic matching or LLM evaluations.

Quickstart Guide: OpenAI Python Integration with LangWatch

Get started with LangWatch quickly using this simple integration guide for OpenAI Python, including setting up autotracking of OpenAI calls

1. Install LangWatch Library:

   pip install langwatch
   

   Copy
2. Trace Your Function:

   import langwatch
   from openai import OpenAI
   
   @langwatch.trace()
   def main():
       client = OpenAI()
       # Your LLM pipeline code here
       langwatch.get_current_trace().autotrack_openai_calls(client)
   

   Copy
3. Set API Key:

   export LANGWATCH_API_KEY='your_api_key_here'
   

   Copy

DSPy Visualizer: A Quick Start

Integrate LangWatch with DSPy to visualize and optimize your prompts and pipelines effectively. Follow these steps to begin:

1. Install LangWatch:

   pip install langwatch
   

   Copy
2. Import and Authenticate:

   import langwatch
   
   langwatch.login()
   

   Copy
3. Initialize LangWatch:

   langwatch.dspy.init(experiment="my-awesome-experiment", optimizer=optimizer)
   compiled_rag = optimizer.compile(RAG(), trainset=trainset)
   

   Copy

Local Development Setup: Run LangWatch on Your Machine

Want to contribute or customize LangWatch? Set it up locally with these steps: Ensure you have Docker and Docker Compose installed.

1. Configure Environment: Duplicate .env.example to .env and update variables.
2. Add API Keys: Include OpenAI or Azure OpenAI keys for LLM guardrails.
3. Set Up Auth0: Create an Auth0 account and update the necessary environment variables(AUTH0_CLIENT_ID, AUTH0_CLIENT_SECRET, AUTH0_ISSUER).
4. Run Docker Compose: Execute docker compose up --build to start LangWatch at http://localhost:3000.

Ready to Dive Deeper? Explore the Documentation

Access detailed guides and resources to maximize your LangWatch experience:

• Introduction to Langwatch
• Getting Started with Langwatch
• Langwatch Integration with OpenAI
• Langwatch Integration with LangChain
• Langwatch Custom REST Integration

Contributing to LangWatch: Join the Community

Your contributions can make LangWatch even better. Check out the Contribution Guidelines to get started.

By implementing LangWatch, you're not just using a tool; you're embracing a platform that fosters clarity, enhances collaboration, and drives the evolution of smarter, more reliable AI applications. Start building AI applications with confidence today!