#

financebench

Here are 31 public repositories matching this topic...

VectifyAI / Mafin2.5-FinanceBench

FinanceBench evaluation of Mafin 2.5 (Powered by PageIndex)

financebench

Updated Oct 20, 2025
Python

avnlp / rag-pipelines

Advanced RAG pipelines for medical (HealthBench, MedCaseReasoning, MetaMedQA, PubMedQA) and financial (FinanceBench, Earnings Calls) QA. LangGraph orchestration + BAML structructed generation, Milvus Hybrid search (Dense + BM25 + RRF), three-layer Metadata Enrichment, Contextual AI instruction-following reranker, and DeepEval evaluation.

pubmed unstructured rag baml milvus earnings-calls contextual-ai llm langgraph rag-pipeline agentic-rag deepeval financebench healthbench

Updated Nov 3, 2026
Python

Rishabhmannu / financebench-rag-agent

Multi-agent LangGraph RAG for financial Q&A — 72.7% on FinanceBench under κ=0.932 calibrated judge. RBAC at the vector layer, multi-party HITL on high-stakes answers, self-hosted LLM observability. pip install financebench-rag-agent

evaluation multi-agent-systems rag qdrant-vector-database retrieval-augmented-generation langfuse langgraph agentic-rag ragas-evaluation financebench

Updated Jun 9, 2026
Python

Zhonghui-li / sec-filing-agent

Vouch — finance-grade tool-calling agent over SEC filings for any US public company: exact XBRL numbers (never LLM arithmetic), every claim cited, abstains instead of fabricating. 92% coverage with zero hallucinations on FinanceBench, 97% of narrative answers grounded, behind a paired per-case regression gate. LangGraph + pgvector on Cloud Run.

agent finance fintech xbrl sec-filings rag llm pgvector llm-evaluation langgraph tool-calling financebench

Updated Oct 19, 2026
Python

ManojDevarajulu / Ledger-AI

Audit-grade financial RAG engine over SEC 10-K filings featuring table-aware chunking, hybrid BM25 + dense search, and deterministic cell verification via sandboxed Python AST.

python bm25 sec-edgar table-extraction rag fastapi financebench llm-verification financial-rag

Updated Oct 26, 2026
Python

adaboranyilmaz / rag-groundedness-eval

Measuring RAG Groundedness on Financial Filings

kubernetes information-retrieval dvc faiss rag mlflow fastapi trustworthy-machine-learning qdrant retrieval-augmented-generation llm-evaluation hallucination-detection llm-as-a-judge financebench groundedness

Updated Sep 29, 2026
Python

cellig / financebench-rag-evaluation

FinanceBench financial-report RAG with FAISS, Flask, Ragas, and financial numeric accuracy evaluation

flask faiss rag llm-evaluation ragas financebench financial-qa

Updated Aug 30, 2026
Python

Abood991B / vectracite-ai

Enterprise Multimodal Document Intelligence, Hybrid RAG (Qdrant + BM25 RRF), Cross-Encoder Reranking & QLoRA/vLLM Engine

kubernetes redis rag mlops fastapi hybrid-search qdrant vllm qlora financebench

Updated Nov 3, 2026
Python

bharathkumardev1 / finance-rag-eval

Page-level retrieval on FinanceBench's SEC filings: a company and period filter and statement routing take held-out Hit@10 from 0.14 to 0.71. BM25 vs MiniLM, load test, browser port.

python information-retrieval bm25 sec-filings rag fastapi financebench retrieval-evaluation

Updated Nov 10, 2026
Python

captof-omuamua / docs-rag-assistant

Chat with your documents: page-cited answers and refusals, fully local on one GPU. 61.8% on FinanceBench (GPT-4-Turbo RAG baseline: 19%).

rag fastapi document-ai qdrant llm vllm retrieval-augmented-generation financebench

Updated Oct 29, 2026
Python

SebAustin / production-rag-eval

Production RAG pipeline on FinanceBench: hybrid BM25+dense+RRF+Cohere Rerank, Anthropic Contextual Retrieval, conformal abstention (Yadkori 2024), Citations API grounding, triple eval (RAGAS+HHEM+DeepEval). Every eval number is reproducible and CI-gated.

python retrieval cohere conformal-prediction rag anthropic evals ragas financebench citations-api

Updated Jun 16, 2026
Python

irenakochtov / RAG-FinanceBench-

a deliberately challenging benchmark built around real financial filings, where retrieving the right evidence is just as important as generating the answer.

python ai embeddings faiss rag huggingface vector-database llm langchain financebench

Updated Jul 27, 2026
Jupyter Notebook

ahmedazaz / enterprise-rag

Local RAG over SEC financial reports, evaluated on FinanceBench: 75.3% top-10 page retrieval, LLM-as-judge answer grading, all experiments measured.

faiss rag streamlit llm ollama llm-as-a-judge financebench

Updated Nov 12, 2026
Python

f20230125-sudo / footnote

A filings analyst that cites the page behind every figure, scored on FinanceBench. Retrieval measured; model runs next.

python mcp bm25 ai-agents rag llm-evaluation financebench

Updated Nov 9, 2026
Python

Ven-Z8 / contextiq

Grounded Q&A over financial filings (10-Ks/10-Qs): routes to the right document, decomposes the question, retrieves, and answers with page citations — or abstains when unsure. Simple-agentic RAG; FinanceBench hardest subset: 0.71 corpus / ~0.79 per-doc vs 0.19 naive.

python finance minimax rag qdrant llm document-qa retrieval-augmented-generation agentic-rag financebench

Updated Jul 18, 2026
Python

andresjmnz92-jpg / pdf-parsing-retrieval

Does a table-aware PDF parser retrieve better? Measured on FinanceBench: 53,901 pages, 150 questions, and a difference that did not survive a paired test.

python retrieval evaluation pdf-parsing rag docling financebench

Updated Aug 10, 2026
Python

bloch-ai-engineering / fin-rag

RAG for financial QA on FinanceBench: FAISS + bge embeddings + Llama-3.3-70B, evaluated on correctness, RAGAS faithfulness and hit-rate

faiss rag langchain retrieval-augmented-generation llm-evaluation ragas financebench

Updated Nov 12, 2026
Jupyter Notebook

chensirui3 / FIling-agent

Research agent that answers analyst questions from SEC filings using the XBRL API, filing search and a calculator. Graded on FinanceBench and on a contamination-free benchmark of post-cutoff filings, with stage-level error analysis and an MCP server.

financial-analysis xbrl sec-edgar rag tool-use llm-evaluation llm-agents model-context-protocol mcp-server financebench

Updated Nov 13, 2026
Python

Nikelroid / footnote-rag

RAG assistant for SEC filings that cites its pages. Evaluated on 150 FinanceBench questions and stress-tested on amended filings, wrong-company pages and prompt injection, across three models. LangChain, LangGraph, Chroma, FastAPI, MCP, Google ADK.

mcp sec-filings rag fastapi langchain prompt-injection chromadb retrieval-augmented-generation llm-evaluation langgraph financebench google-adk

Updated Nov 12, 2026
Python

Yukseltt / finans_rag

SEC dosyaları üzerinde ölçülmüş finans RAG hattı: e5 + belge yönlendirme + reranker + Gemini. Ön kayıtlı deneyler, kilitli test, olumsuz sonuçlar dahil.

finance retrieval gemini reranking sec-filings rag pre-registration llm-evaluation financebench

Updated Oct 6, 2026
Python

Add this topic to your repo

To associate your repository with the financebench topic, visit your repo's landing page and select "manage topics."