Our Take
Both moves are tactical integrations of existing AI infrastructure into knowledge workflows, not capability breakthroughs; the real test is whether law firms actually use these connections instead of building custom layers.
Why it matters
Legal teams face acute friction: subject matter experts won't spend time tagging documents, and AI vendors require grounded knowledge to deliver accurate results. These products address the plumbing problem, but only if adoption scales beyond early adopters.
Do this week
Knowledge management teams: audit your current AI tool stack (Copilot, Claude, Harvey, etc.) this week and map which Lexsoft T3 or Tiger Eye Blueprint connectors match your workflow before requesting budget for either platform.
Two legal AI knowledge platforms announce vendor integrations
Lexsoft Systems announced on May 18 that its knowledge management system, T3, is now accessible via the Model Context Protocol (MCP). The integration allows law firms and corporate legal departments to connect T3 with MCP-compatible platforms including Microsoft Copilot, Claude, and Gemini, as well as legal-specific AI tools like Harvey. Lexsoft also released a new OpenAI-based vectorised Indexer for T3 that performs semantic search, distinguishing conceptual similarity (contract vs. agreement) rather than exact word matching alone.
Separately, Tiger Eye launched an AI Curation Assistant for its knowledge management platform, Blueprint. The feature uses Azure OpenAI to auto-suggest metadata, tags, and taxonomy fields based on document content, reducing manual enrichment work before documents are contributed to central repositories.
Knowledge engagement remains the bottleneck in legal AI
Both announcements target the same operational friction: law firms struggle to get busy partners and subject matter experts to contribute knowledge to shared systems. Manual tagging is time-consuming, and unstructured documents reduce the accuracy of AI-driven document retrieval.
The MCP approach lowers switching costs by treating T3 as a data source pluggable into any MCP-compatible AI platform. Rather than forcing law firms to adopt a single vendor's AI stack, T3 becomes a backend layer that AI tools call on. Tiger Eye's curation assistant targets the upstream problem: making contribution less friction-intensive so lawyers actually participate in knowledge-building.
The semantic search capability in Lexsoft's new indexer addresses a known failure mode of keyword-only retrieval: missing relevant documents because lawyers use different terminology than the original author. Context-aware search, however, requires a trained embedding model and sufficient volume of indexed documents to be effective. Early-stage implementations often return false positives.
Three questions before committing to either platform
Does your firm already use one of the target AI vendors? Lexsoft's MCP integration only creates value if you have an active workflow in Copilot, Claude, or Harvey. If your firm standardizes on a different platform, the integration is irrelevant.
Is knowledge contribution actually the constraint, or is it retrieval quality? Tiger Eye's feature solves tagging friction. If your problem is that subject matter experts won't share knowledge at all (a cultural issue, not a UX issue), AI-assisted metadata won't change behavior. Conversely, if documents are being shared but getting lost in search, Lexsoft's semantic indexer is the lever to pull.
Who owns the semantic model? Neither announcement clarifies whether the Lexsoft indexer or Tiger Eye's tagger remain vendor-controlled or whether embedding models can be updated or replaced as legal LLMs improve. Semantic search quality degrades over time as baseline models age. Confirm model ownership and update frequency before rollout.