Consolidate Datasets, Design MongoDB -- 2

I maintain several large, structured datasets spread across MySQL, PostgreSQL and SQL Server instances. Only portions of the existing schemas are documented, so the very first step will be to inspect each system, reverse-engineer any missing metadata and confirm table relationships.

Once the landscape is clear I need all records extracted, cleaned, normalised and deduplicated, then merged into a single, well-organised MongoDB database. Your work should include a carefully planned document model that supports growth, fast look-ups and straightforward analytics. Proper indexing, sharding strategy (where appropriate) and query optimisation are critical because future workloads will be heavy and latency-sensitive.

Please deliver:
• An optimised MongoDB schema and architecture diagram
• Fully migrated data, validated for completeness and integrity
• Performance benchmarks demonstrating expected scalability
• Clear, developer-friendly documentation of the structure and migration process

If you have handled scientific or biomedical datasets before, let me know—that context would be a bonus. A clean, reproducible pipeline and well-commented code will set the foundation for ongoing collaboration on analytics and feature development after this milestone is met.