Evidence · first-party tested
On the two-column resume, both certification entries lose the issuer field entirely; the output keeps only name and year for 2 certifications even though issuer text is present in the source.
⚠ StruggledLlamaParse →
A two-column, sidebar-style resume for Priya Sharma designed to test whether parsers can preserve reading order and correctly extract content from split layouts.
The observation
On the two-column resume, both certification entries lose the issuer field entirely; the output keeps only name and year for 2 certifications even though issuer text is present in the source.
Criterion: Accuracy · Scenario: Multi-column sidebar resume — Priya Sharma (group: resume-parsing-benchmark)
Query this
get_evidence({
tool: "llamaparse",
scenario: "resume-parsing-benchmark"
})MCP · mcp.aidemos.com/api/mcp
Free with attribution.
Free with attribution.
Same input, other tools
Real outputs, no retouching · every cell queryable via API & MCP · aidemos.com