← All work

Industry work

Financial language systems

Named-entity recognition, table extraction and automatic evaluation for financial question answering.

2023Role period

The work

Built an automatic evaluation framework for a financial chatbot, reducing required human evaluation time by 67%. Trained Flan-T5 XXL for financial NER and built table-extraction techniques; reported QA accuracy exceeded 90% over 401K documents.

Context

Financial language systems and automatic evaluation.

Scope

Industry outcomes are drawn from my resume. Client source code and implementation details are not publicly shared.