MeshView — Zero-copy federated data layer with entity resolution
8/10
Demand Score
Teams can’t answer cross-system questions or build features without manual CSV extracts and brittle point-to-point joins.
8/10
Blue Ocean
Competition Level
$2k-12k
Price/Month
Predicted customer spend
14 days
Time to MVP
Difficulty: Hard
The Problem
A virtualized data mesh that exposes unified GraphQL and SQL endpoints over SaaS APIs, data warehouses, and operational DBs. Defines canonical "Blueprint Entities" that stitch siloed datasets using pr
🔗 Validated by Real User Complaints
This problem has been verified through 3 real user complaints:
Competitor Landscape
- Denodo
- Starburst/Trino
- Hasura
- Apollo Federation
- Presto
- Snowflake + external tables
Must-Have Features for MVP
Blueprint Entities with policy-defined join keys
Probabilistic entity resolution with feedback loop
Adaptive caching with TTL and invalidation by webhook
Live schema registry and federation planner
Row/column/data product level access controls
Audit trails and data lineage across sources
Visual view composer and one-click materialization to warehouse
⚠️ Potential Challenges
- Performance over high-latency SaaS APIs
- Accurate entity resolution across noisy keys
- Row- and column-level security enforcement
- Query pushdown and caching correctness
Risk Level: Moderate
🎯 Keys to Success
- Reduce manual exports by >60%
- P95 federated query latency under 2s for cached views
- Accurate entity linking >98% on validated samples
- 3+ cross-silo use cases live in first 30 days
- Zero sensitive data exfiltration incidents
Ready to Build This?
This hard-difficulty project could be your next micro-SaaS success.