How Saga Uses DataStax Astra DB and Microsoft Azure to Deliver Contextual Legal Intelligence at Scale
Up to 80% faster document review
Fastest Hybrid Search in the market
Seamless hybrid search delivering high-precision legal insights at scale
Up to 80% faster document review
Fastest Hybrid Search in the market
Seamless hybrid search delivering high-precision legal insights at scale
Saga is a legal AI company dedicated to enhancing access to justice and enabling legal professionals, such as law firms, legal aid organizations, and courts, to work more efficiently and accurately through artificial intelligence. Designed as a human-in-the-loop-centric platform, Saga integrates AI deeply into legal workflows while maintaining transparency and control. Saga relies on Astra DB for scalable vector search and Microsoft Azure for secure, compliant infrastructure to power its contextual AI capabilities.
Challenge
Legal professionals operate in highly complex, context-rich environments where they must evaluate various information sources, such as emails, contracts, case law, timelines, and client communications, to form a comprehensive legal picture. Traditional tools such as static document repositories and basic keyword searches fall short in supporting this work's nuanced and dynamic nature. Saga required a solution to ingest and unify legal data in multiple formats and from various sources, while enabling contextual retrieval through retrieval-augmented generation (RAG) without jeopardizing sensitive data. The system supports hybrid search across keyword and semantic vectors, integrates seamlessly with various best-in-class large language models (LLMs), and meets strict European data protection standards for commercial use. Most importantly, Saga’s solution must be intuitive and trustworthy for legal professionals who might not have prior experience with AI tools.
Solution
Saga chose Astra DB as the core vector database for storing and retrieving context in legal workflows. Astra DB enables high-performance hybrid search, powering applications that easily summarize case files, draft contracts, analyze arguments, and extract insights from diverse legal documents. Saga uses Astra DB to:
- Ingest multimodal legal content—emails, court rulings, contracts, and more—into a vector-native structure.
- Execute hybrid queries that combine traditional keyword matching with vector similarity for more precise results.
- Support RAG pipelines where retrieved documents are assembled into context-aware prompts for LLMs.
- Optimize retrieval relevance through re-ranking techniques, custom embeddings, and review agents.
“Astra DB is the best vector database out there. It’s easy to implement and flexible enough to adapt to any legal use case we throw at it,” said Remco Visser, Co-founder and Legal AI Engineer, Saga.
Microsoft Azure Integration
Saga runs its entire stack on Microsoft Azure, leveraging European data centers to ensure compliance with GDPR and legal industry privacy requirements. Both the Saga platform and Astra DB are hosted in the same Azure environment, enabling secure, low-latency private networking.
Benefits of the Azure integration include:
- Data privacy and security aligned with legal standards, including support for professional secrecy
- Infrastructure compatibility with Microsoft tools that legal firms already use (Outlook, Teams, SharePoint)
- LLM flexibility via Azure AI Foundry, which enables Saga to use models from OpenAI, Mistral, and other providers
- Rapid deployment with infrastructure templates that reduce overhead and enable faster customer onboarding
- Support for sensitive content through custom filtering policies provided by Microsoft
“The perfect combination of Saga, Microsoft, and DataStax gives us security, flexibility, and speed. Azure makes deployment seamless, and customers trust the Microsoft environment,” said Visser.
Technical architecture
Saga’s legal AI platform is hosted entirely on Azure, with Astra DB deployed in the same regional cloud environment. This unified architecture ensures enterprise-grade security and simplifies integration across the stack. The platform ingests legal documents, such as contracts, email threads, and jurisprudence, into Astra DB, transforming them into vector embeddings for efficient contextual retrieval. Saga orchestrates end-to-end RAG workflows that retrieve documents from Astra DB, construct prompts, and route them to LLMs deployed securely via Azure AI Foundry.. This system supports multiple model providers, allowing Saga to remain LLM-agnostic and flexible.
Search results are optimized using re-ranking, custom embedding models, and agentic review layers—all built in collaboration with DataStax. With Astra DB Hybrid Search and Azure’s compliance-focused cloud, Saga delivers fast, relevant, and secure AI-driven legal assistance to end users.
“Without Astra DB, we couldn’t provide our users the kind of large-context legal AI assistance they need,” said Knut-Magnar Aanestad, Co-founder and Legal Process Engineer, Saga.
Results
- Up to 80% faster document review
- The accuracy of results for legal-specific training, onboarding, and transparency drives high user adoption
- The fastest Hybrid Search in the market
- Rapid feature development with two-week development sprints using AI-assisted coding
- Regulatory compliance across Europe through fully EU-hosted Microsoft Azure infrastructure
- Seamless hybrid search delivering high-precision legal insights at scale
Roadmap
Saga’s future roadmap combines continued technical innovation with enhanced engagement within the legal community.
On the technical front, the team is expanding integrations with legal research tools and document management systems to bring even more content into Astra DB. Using Langflow, Saga is building task-specific agents that automate end-to-end legal workflows—from contract review to regulatory compliance checks. These capabilities are first rolled out internally to Saga’s legal engineers, then co-developed with client law firms to match real-world needs.
On the user engagement side, Saga is launching a community learning platform featuring webinars, prompt templates, and recorded demos. This supports its Saga Amplify program, which provides AI adoption training for legal professionals and promotes AI's ethical and practical use in law.
“Our mission is to support lawyers, not replace them. That means building AI they can trust—focusing on relevance, accuracy, and transparency above all else,” said Aanestad.
By combining Astra DB for vector search and Azure for secure, scalable infrastructure, Saga has built a contextual AI platform that meets the real-world needs of legal professionals. Its lawyer-first design philosophy, deep AI integration, and transparent approach to adoption set Saga apart in a rapidly evolving field. With support from DataStax and Microsoft, Saga is transforming legal work, making it faster, more accurate, and more accessible than ever before.
