Process mining for the classroom

Mine your process.
Fix what slows you down.

Free for students, faculty, and researchers. Sign up with your institutional email, upload an event log, and run the algorithms you studied in class — Alpha Miner, Inductive Miner, IMf, DFG, Heuristic Miner, Split Miner, or the Genetic Miner — plus PHM, ClikFlow's own production algorithm, with every result reproducible and auditable for your methodology section.

Create Free Academic Account Sign In
CSV
XES
OCEL 2.0
SAP
8 discovery algorithms
Token Replay — Live
Receive Order Credit Check ⚠ Pack Items Send Invoice Ship & Close
Token
Activity
Bottleneck
No installation required
Results in seconds
LGPD compliant
CSV · XES · OCEL 2.0 · SAP
Free for academia
Three steps
From raw data to process insights
1
Upload your event log
Drop a CSV, XES, or OCEL 2.0 file, or connect your SAP system. ClikFlow Academy auto-detects columns and maps Case ID, Activity, and Timestamp — no configuration needed.
2
Discover the real process
Choose between Alpha Miner, Inductive Miner, IMf, DFG, Heuristic Miner, Split Miner, the Genetic Miner, or PHM — run two side by side to compare against the algorithm you studied in class, with the exact parameters recorded for your write-up.
3
Find & fix bottlenecks
Inspect the interactive process map, replay tokens through your cases, rank bottlenecks by wait time, and compare variants. Save every finding as an investigation.
Token replay
Watch how process replay works
Each orange dot is a case token flowing through discovered activities. The blinking red transition is where cases pile up — your bottleneck.
ClikFlow — Token Replay Simulation · Purchase-to-Pay Process Live
Receive Order Credit Check ⚠ Prepare Package Generate Invoice Ship & Close avg wait: 4.2 days
Place
Activity
Token (case)
Bottleneck
Arc (flow)
The interface
Everything in one place
localhost:8000/app
Process Map
Variants
Bottlenecks
Conformance
Receive Order 452 cases · 0.3d avg Credit Check ⚠ 4.2d avg · P95: 9d Pack Items 0.8d avg Send Invoice 0.4d avg Ship 1.1d avg
Quality metrics
Fitness 0.94
Precision 0.87
Generalization 0.91
Top bottlenecks
Credit Check 4.2d
Ship Order 1.1d
Pack Items 0.8d
Send Invoice 0.4d
Log summary
Cases 452
Events 3 841
Activities 5
Variants 12
What you can do
Everything your process needs
📂
Upload & Preview
Drop in a CSV or XES event log. Auto-detect delimiters, preview all rows, and tag your Case ID, Activity, and Timestamp columns instantly.
🗺️
Process Discovery
Discover the real process model from data — not from what people think they do. See activity flows, parallel paths, and confidence scores in an interactive map.
⏱️
Bottleneck Analysis
Ranked bottleneck table with P50/P95 wait times and impact scores — so you know exactly which activities are costing you the most time.
💻
Bring Your Own Algorithm
Implement a discovery algorithm from a paper — or your thesis's own contribution — in Python, and run it against your event log in an isolated sandbox, right next to the built-in miners.
🔍
Investigations
Save every analysis as an investigation. Replay past process maps, run two algorithms side by side to compare metrics, and track how improvements evolve over time.
🔀
Variant Explorer
Drill into process variants — see which paths cases actually follow, highlight deviations on the map, and spot the patterns behind performance gaps.
🧾
Reproducible & Auditable
Every result records the exact parameters used, a hash of the input dataset, and the software version — export it as JSON or a print-ready report for your methodology section.
🔌
SAP Connector
Connect directly to SAP via RFC or OData. Pull Purchase-to-Pay, Order-to-Cash, or Finance event data without writing a single line of SQL.
ClikFlow's own algorithm
Meet PHM — the Precision Hybrid Miner

The same discovery algorithm behind ClikFlow's commercial product, free to run on Academy. Instead of relying on a single dependency measure like Alpha or Heuristic Miner, PHM blends three signals into one hybrid score per activity pair — then checks its own work with real token-based replay, the same methodology used to grade every other algorithm on this page, so the comparison is apples-to-apples.

🧮
Hybrid dependency score
Combines Heuristic-style dependency strength, statistical confidence, and relative-frequency scoring into a single weighted score — more robust than any one metric alone on noisy, real-world logs.
🔀
Real XOR-split detection
Identifies mutually-exclusive branches directly from trace co-occurrence, not just edge frequency — surfacing decision points that simpler dependency-graph miners tend to blur together.
🎚️
Tunable, not black-box
Three explicit thresholds — dependency, frequency, and noise — let you trade off fitness against precision on purpose, with every value recorded alongside the result for your methodology section.
Try PHM on your own log Run it side by side with Alpha, Inductive, or Heuristic Miner to see the difference
For Students & Researchers

Learn process mining hands-on — for free

ClikFlow Academy is a dedicated, isolated environment for process mining courses — sign up with your institutional email (.edu, .ac.xx, or a recognized university domain) and tell us your university or college. Upload the same CSV, XES, or OCEL 2.0 logs from your assignments, run the algorithms you studied in class, and see the results in a real interactive interface — not just theory on a slide.

No installation. No Python setup. No configuration files. Just upload and discover — with every parameter, dataset hash, and software version recorded so your results hold up in a methodology section.

Free account Institutional email required CSV · XES · OCEL 2.0 Export results for papers 8 discovery algorithms Conformance checking Token replay Reproducible & auditable
α
Alpha Miner
The foundational discovery algorithm (van der Aalst, 2004). Derives a Petri net directly from footprint relations — the classic starting point in every process mining course.
IM
Inductive Miner & IMf
The current gold standard in process discovery research, guaranteeing sound models. IMf adds a noise threshold to compare fitness/precision trade-offs on real-world logs.
🕸️
Directly-Follows Graph
The "X follows Y" graph every other miner is built from — the pedagogical starting point for teaching process discovery from first principles.
HM
Heuristic Miner
Weighs directly-follows relations by frequency instead of a strict present/absent footprint — the classic approach behind ProM and Disco, more robust to noise and short loops.
🧩
Split Miner
Discovers a BPMN model directly (Augusto et al., 2019) — the modern state-of-the-art baseline for simple, readable models. Download the result as real BPMN 2.0 XML.
🧬
Genetic Miner
Evolutionary (NEAT-based) process discovery. Evolves a population of candidate process models over generations to optimize fitness, precision, and simplicity together.
🎯
PHM ClikFlow's own
Precision Hybrid Miner — blends dependency strength, statistical confidence, and frequency into one score, with real XOR-split detection from trace co-occurrence. The same algorithm behind ClikFlow's commercial product.

Ready to see your real process?
Start mining in seconds.

Upload your first event log for free with your institutional email. No credit card, no installation, no SQL.

Create Free Academic Account Sign In

Free for verified students, faculty, and academic researchers · LGPD compliant