How pranjasidhwani1/latency_priortized_smartmapper_using_ai works
Smartmapper maps financial-file headers (and, where applicable, categorical values) to standardized fields, using company history, fuzzy matching, optional embeddings, and LLM suggestions. The principal workflow enters through the company-level wrapper, loads a file, delegates mapping to Smartmapper, optionally adds QC derivability results, and returns a response. Separate runtime capabilities include raw-file parsing and feedback-based history updates. Source excerpts show the wrapper and QC calls; some matching internals are documented in the README but not fully sampled.
Main entry points: smartmapper.py Smartmapper · smartmap_company_level.py Company mapping · get_derivable_unmapped.py QC derivability · update_training_data_company_level.py Feedback update
Components 15
Mapping workflow
- Company mapping
smartmap_company_level.py - File loading
smartmapper.py - Smartmapper
smartmapper.py - Mapping result
Matching methods
- History matching
smartmapper.py - Fuzzy matching
smartmapper.py - Embedding matching
bert_model.py - LLM prompt
prompt.py
QC and file parsing
- QC derivability
get_derivable_unmapped.py - Central config DB
- Raw-file parser
parse_raw_data.py
Mapping history
- Company mapping history
smartmapper.py - Feedback update
update_training_data_company_level.py
Other components
- Data user
- AWS Bedrock
Connections 16
- Data user → Company mapping: submits file
README.md - Company mapping → File loading: loads file
smartmap_company_level.py - Company mapping → Smartmapper: runs mapping
smartmap_company_level.py - Smartmapper → History matching: checks history
README.md - Smartmapper → Fuzzy matching: matches terms
README.md - Smartmapper → Embedding matching: matches semantically
README.md - Smartmapper → LLM prompt: builds prompt
smartmapper.py - Smartmapper → AWS Bedrock: requests suggestions
README.md - Smartmapper → Mapping result: returns mappings
README.md - Company mapping → QC derivability: checks unmapped fields
smartmap_company_level.py - QC derivability → Central config DB: queries config
get_derivable_unmapped.py - QC derivability → Mapping result: adds QC results
- Data user → Raw-file parser: parses file
parse_raw_data.py - Feedback update → Smartmapper: updates history
update_training_data_company_level.py - Smartmapper → Company mapping history: reads and writes
README.md - Feedback update → Company mapping history: exports updates
update_training_data_company_level.py