自然语言驱动的自动化 SQL 对账系统 — 基于 HelloAgents 多 Agent 框架 · ReAct 推理 · NL2SQL
-
Updated
Jun 19, 2026 - Python
自然语言驱动的自动化 SQL 对账系统 — 基于 HelloAgents 多 Agent 框架 · ReAct 推理 · NL2SQL
[ARCHIVED] A story-driven AI system demonstrating data reconciliation through a cyberpunk animated universe. — SSL v3.0
Task-post consistency audit CLI for databases and lakehouse tables (Iceberg/Trino/JDBC) - pinpoints which rows differ and why
Local-first CLI for semantic data reconciliation with YAML rules and JSON reports.
Digital Scriptorium OpenRefine documentation and JSON recipes for data reconciliation
Quality Engineering portfolio focused on Data Quality, ETL testing, PostgreSQL, SQL, reconciliation, large-volume validation and CI/CD.
Initial public source, qualification scripts, compatibility documentation, and reproducibility checks for FastDiff v0.2.1.
Pre-production. Deterministic evidence-domain core for California onsite nonpotable water reuse (OTNWS) programs under 22 CCR 60606: governed CSV normalization, cross-source reconciliation, frozen report drafts, unsigned receipts. Synthetic fixtures only. Shares its reconciliation layer with a private sibling repo; canonical copy undecided.
Public data pipeline with FastAPI, Celery, and PostgreSQL for collecting, normalizing (Bronze/Silver/Gold), and reconciling financial data from Jundiaí/SP.
Human-in-the-loop Applied AI case study for multi-source crypto/VDA reconciliation, traceability, exception handling, and synthetic validation.
55 published n8n workflow templates: fraud screening, CSV reconciliation, deadline calculation, approval gates, SLA audits, and grounded RAG drafting. Deterministic rules, audit logs, human review before anything sends.
Reconcile company fundamentals from EODHD and FMP into one schema and see exactly where the two vendors disagree. Hand-curated 71-field mapping, configurable vendor priority, and a Streamlit dashboard.
Industrial data reconciliation platform using Lagrange Multipliers and statistical validation
🍺 End-to-end B2B EDI Supply Chain Transformation Portfolio in Python & SQL. Features ANSI X12 (850, 856, 810) payload reconciliation against SAP staging, real-time exception monitoring with SLA tracking, and partner migration ETL pipelines with Colab support.
Reproducible ESG KPI extraction and reconciliation pipeline for PDF disclosures, combining tables, regex, NLP and optional LLMs with provenance-aware evidence and controlled synthetic benchmarking.
Declarative data-reconciliation engine in Go — exact→tolerance→fuzzy→ML cascade matching, audit trail, React workbench. Self-hosted & deterministic.
Enterprise data reconciliation and risk reporting platform for validating transactional data, detecting exceptions, reconciling source systems, and generating automated control reports.
Three-tier migration reconciler: row counts, order-insensitive column checksums, and exact DECIMAL business aggregates, with checksum-bisection drill-down to the exact divergent rows. Clean 5M-row pair verified in 0.55s; 1,000 divergent rows localized in 18 grouped probes without transferring tables.
Deterministic source-to-target data reconciliation: quantitative and qualitative checks, a browser UI, reports, notifications, and an MCP server. Apache-2.0.
[Moved to warehouse-ai-lab/claude-ops-toolkit] AI-assisted ERP+WMS ops toolkit: Claude Code diagnostic skills + always-on Windows Task Scheduler monitors with Slack alerts (sanitized reference).
To associate your repository with the data-reconciliation topic, visit your repo's landing page and select "manage topics."