v1.1.0

Algorithmic Process Mining v1.1.0

A production-shaped process mining and conformance checking pipeline for logistics event logs (Order-to-Cash / Procure-to-Pay) — discovery, conformance checking (as-is vs. to-be), bottleneck/rework analytics, and object-centric (OCEL 2.0) support, driven by a CLI or an interactive Streamlit dashboard.

Highlights

• Discovery — Inductive Miner (im/imf/imd variants) via pm4py
• Conformance checking — token-based replay + alignments against a normative SOP model, with real fitness scores
• Root-cause analytics — bottleneck detection, rework loops, trace variant analysis
• Object-centric process mining (OCEL 2.0) — now reachable end-to-end via the new ocel-summary CLI command
• On-disk Parquet caching — opt-in --cache flag skips re-ingestion/re-transformation on repeat runs against the same source file
• Config-driven input allowlisting — optional data.allowed_root confines --input paths, ready for non-local deployment scenarios

Quality & security

• 142 tests passing, 97% coverage on src/
• ruff, ruff format, mypy --strict, bandit, and pip-audit all clean in CI (Python 3.11/3.12)
• Full security & vulnerability audit in docs/AUDIT_REPORT.md: dependency CVE research, supply-chain legitimacy verification, static code-integrity review, and authorized adversarial testing (including a confirmed-and-fixed CSV formula injection vulnerability)

CLI

generate --process {o2c,p2p} --cases N [--noise F] [--seed N] --output PATH
run --input PATH --output PATH [--anonymize] [--salt S] [--config PATH] [--cache]
discover --input PATH --output PATH [--variant {im,imf,imd}] [--noise-threshold F] [--cache]
conformance --input PATH --output PATH [--model PATH] [--cache]
ocel-summary --input PATH --output PATH

Or launch the dashboard: streamlit run src/viz/dashboard.py

See README.md for the full quickstart, configuration, and Docker instructions.