Azure AI Document Intelligence client library for Python - a cloud service that uses machine learning to analyze text and structured data from documents
76
{
"context": "Evaluates whether the solution uses azure-ai-documentintelligence APIs to retrieve stored analysis outputs as searchable PDFs and figure images, and cleans up stored results safely. Checks focus on correct client calls rather than generic HTTP or filesystem logic.",
"type": "weighted_checklist",
"checklist": [
{
"name": "PDF retrieval",
"description": "Uses DocumentIntelligenceClient.get_analyze_result_pdf (or async equivalent) with the provided result_id to stream the searchable PDF output before writing it, rather than reimplementing the request or pulling non-PDF content.",
"max_score": 25
},
{
"name": "Figure fetch",
"description": "Fetches each requested figure via get_analyze_result_figure (sync or async) using the correct figure index and honors the requested image_format when saving the bytes.",
"max_score": 25
},
{
"name": "Figure discovery",
"description": "When figure_indexes is empty/None, first calls get_analyze_result to read available figure identifiers and iterates over that set instead of assuming indexes or counts.",
"max_score": 20
},
{
"name": "Conditional delete",
"description": "Invokes delete_analyze_result only when delete_after is true and after all downloads succeed, using the same result_id and avoiding redundant delete calls.",
"max_score": 20
},
{
"name": "Skip delete on failure",
"description": "If PDF or figure retrieval throws, does not call delete_analyze_result and propagates the failure, rather than cleaning up partial runs.",
"max_score": 10
}
]
}Install with Tessl CLI
npx tessl i tessl/pypi-azure-ai-documentintelligencedocs
evals
scenario-1
scenario-2
scenario-3
scenario-4
scenario-5
scenario-6
scenario-7
scenario-8
scenario-9
scenario-10