Legal Discovery Hybrid Search Database Selection Matrix

Role: Senior Legal Engineering Database Specialist focused on regulatory compliance and search infrastructure.

Context

• Practice Area: {{legal_practice_area}}
• Corpus Size: {{total_document_volume}}
• Query Latency SLA: {{retrieval_latency_sla}}
• Residency Requirements: {{jurisdictional_residency_rules}}
• Search Workload Mix: {{search_modality_mix}}
• Base Infrastructure: {{existing_infrastructure}}

Task

Produce a technical evaluation matrix contrasting candidate database configurations for a legal discovery platform to determine the optimal engine for mixed vector, full-text, and metadata queries.

Method

1. Deconstruct the workload mix from {{search_modality_mix}} into exact keyword, faceted metadata, and vector similarity retrieval requirements.
2. Assess the scalability requirements for storing and indexing {{total_document_volume}} within {{existing_infrastructure}}.
3. Frame database candidates across three categories: dedicated vector database, integrated relational vector store, and enterprise search cluster.
4. Evaluate index rebuild times and query execution times against {{retrieval_latency_sla}}.
5. Score candidate databases on strict adherence to {{jurisdictional_residency_rules}} including field-level encryption and immutable audit trails.
6. Analyze operational overhead specific to litigation holds and rapid evidence ingestion in {{legal_practice_area}}.
7. Compile findings into a comparison matrix with a weighted trade-off analysis.

Constraints

• MUST compare hybrid search capabilities (combining BM25/keyword with dense embeddings).
• MUST evaluate immutability, audit logging, and {{jurisdictional_residency_rules}} for each candidate.
• Do not recommend proprietary hosted solutions that violate {{existing_infrastructure}} deployment constraints.
• Candidate solutions MUST be technically viable for {{total_document_volume}}.

Output format

1. Scope Statement (1 brief paragraph defining workload and scale parameters).
2. Technical Evaluation Matrix (Markdown table with columns: Candidate Engine, Hybrid Search Quality, Indexing Speed, {{jurisdictional_residency_rules}} Compliance, Operational Complexity, Fit Score [1-10]).
3. Trade-off Analysis (3 bullet points highlighting primary trade-offs: latency vs. cost, schema flexibility vs. strict auditability).
4. Definitive Architecture Selection (1 short paragraph with final recommendation).

Self-review

• Ensure the matrix explicitly accounts for {{search_modality_mix}} requirements.
• Confirm that {{retrieval_latency_sla}} is specifically evaluated in the engine review.
• Check that every candidate supports required residency and compliance boundaries.