In government contracting, the instinct is always the same: one more source. One more portal. One more archive. One more spreadsheet maintained by someone who retired three administrations ago. We’re proud to announce Sweetspot’s newest data source: Mesopotamian cuneiform search. Because if your goal is to collect all the government contracting data sources, eventually you have to go back to where public administration really started — clay tablets, grain ledgers, labor records, and procurement notes pressed into wet mud sometime around 2100 BC.
The feature is new. The data is not.
Searching for ancient administrative records in Sweetspot — just like any other data source.
Procurement people have always been procurement people
Long before SAM.gov and FPDS, administrators in Mesopotamia were already tracking deliveries, labor, inventories, and budget allocations. The funniest part? Many of those ancient records are officially classified as Administrative — which means we can say, with a completely straight face, that we’re expanding Sweetspot’s coverage of administrative contracting records by roughly four millennia. That’s what we call deep market intelligence.
What you can search
The records cover what you’d expect from any good procurement archive: vendor names, quantities, delivery locations, time periods, and genre. The genre field, for the record, comes up Administrative a lot. Users can run queries like:
- Find grain allocation records from the Ur III period
- Show me the oldest available documentation of budget execution anxiety
- Surface proto-ledger entries that feel spiritually identical to subcontractor tracking
Saved searches are particularly useful for recurring topics like barley, livestock, and labor allocation — the timeless indicators of organizational planning.
Search results with tablet imagery, transliterated text, and metadata — 4,000 years old, fully indexed.
The procurement problem has always been the same: information lives in too many places and the important records are hard to find. Sometimes the source material is a CSV export with broken headers. Sometimes it’s a clay tablet.
Credit where it’s due
The dataset powering this is CuneiML, published by Danlu Chen, Aditi Agarwal, Taylor Berg-Kirkpatrick, and Jacobo Myerston in the Journal of Open Humanities Data. It’s a genuinely impressive piece of work — high-quality tablet imagery, transliteration, and metadata, all structured for computational use. Honestly, better organized than some federal data portals we could name.
The fine print
Your BD team does not need to learn Akkadian. We are not replacing SAM.gov with a clay tablet ingestion pipeline. And we are not yet offering cross-source search between modern federal opportunities and 4,000-year-old administrative records — although that would be an extremely strong differentiator.
Happy April Fool’s Day.