Vexl Systems
LESS OVERHEAD. MORE POSSIBILITY.

Big compute.
Small footprint.

Native compute. Analytical SQL. Process mining. Discover how your operations run, investigate the friction and keep control of the data underneath.

Explore VexlFlow ↗Find your plan →
NEW / VEXLFLOW   Your process-mining alternative to Celonis. Explore 8 client scenarios →
Native execution Memory-conscious design
THE VEXL FAMILY

VexlPy. VexlSQL. VexlFlow. VexlSmart. One systems mindset.

VEXLFLOW / NATIVE PROCESS MINING

Find the friction.
Follow the flow.

Your process-mining alternative to Celonis.

Reconstruct how work actually moves. Find process variants, long handoffs and repeated steps in your event logs—with native analysis and an explorer that stays on your device.

Try the process explorer ↓
01 / DISCOVER

See the paths
behind the process.

Turn case IDs, activities and timestamps into an observed process map and ranked variants.

02 / INVESTIGATE

Make every
handoff visible.

Explore elapsed times, repeated activities and individual case timelines. Find where to ask the next question.

03 / KEEP CONTROL

Native analysis.
Local exploration.

Analyse CSV event logs with the native CLI. Open the resulting HTML report locally, without a web server.

BUILT AROUND THE BUYER’S PROCESS

Start with an operational question.

Eight fictional client scenarios. Choose a process to explore its paths, handoffs and exceptions. These are test datasets, not customer results.

VEXLFLOW / PROCESS EXPLORER

From event log to operational insight.

Fictional cross-border payments dataset. Demonstration only.

0.1.0 / DEVELOPER PREVIEW
BANKING / FICTIONAL CLIENT PROFILE

Payment operations director at a regional bank

Where do repair loops and correspondent handoffs extend the observed payment journey?

    START WITH A REPRESENTATIVE EVENT LOG

    Case identityOne synthetic payment reference

    Source systems to scopePayment hub, screening event export, repair workbench and correspondent acknowledgements.

    Data requirements & preview boundaries +

    LOCAL TO YOUR BROWSER

    CASES ANALYSED96586 accepted events
    MEDIAN CASE DURATION2.2 hAcross the full event log
    PROCESS VARIANTS6Distinct ordered activity sequences
    REPEATED-ACTIVITY CASES10.4%Candidates for review, not proven defects

    Observed process map

    Select a variant to highlight its route. Full-log metrics stay unchanged.

    Process variants

    Case explorer

    CaseVariantDuration

    CSV analysis runs in the native executable. This browser imports and explores JSON reports.
    EVALUATING A CELONIS ALTERNATIVE?

    Start with one process.
    Prove the fit on your own log.

    Use an exported event log to evaluate process discovery and local reporting. Agree on the questions, data quality and success criteria before expanding the deployment.

    IN THE CURRENT PREVIEW

    Case-centric CSV analysis, process maps, ranked variants, elapsed handoffs, case timelines and data-quality observations. Linux x86_64 native binary; single-threaded execution.

    NEXT PRODUCT MILESTONES

    Enterprise connectors, conformance checking, object-centric relationships, multi-user governance, parallel processing and VexlSQL-backed persistence.

    This is a focused developer preview, not full Celonis feature parity. Production readiness, migration scope and recurring pricing are agreed after workload evaluation.

    01 / THE PLATFORM

    Four products.
    One systems mindset.

    Compute, data, process discovery and private intelligence. Choose the product that fits your workload.

    >_ENGINE PREVIEW

    VexlPy

    Keep the way you think.
    Change how your code runs.

    A Python-shaped compiled language and data engine with native execution and explicit parallelism.
    pip install vexlpy
    Target package name · release in preparation
    ≋ENGINE PREVIEW

    VexlSQL

    Serious queries.
    A deliberately compact engine.

    Native analytical SQL with compressed column storage, selective reads and a PG-wire subset in preview.
    vexlsql
    Target binary name · proprietary native engine
    ⑂PROCESS MINING / PREVIEW

    VexlFlow

    See the process.
    Find the friction.

    Native event-log analysis for process maps, variants, elapsed handoffs and case timelines. A focused local alternative for teams evaluating process mining.
    vexlflow analyze
    0.1.0 native preview · Linux x86_64 validated
    ✳ON THE ROADMAP

    VexlSmart

    Intelligence where
    your data lives.

    A roadmap for air-gapped SQL assistance, smart-meter VEE and industrial anomaly intelligence.
    vexlsmart
    Planned module and API
    02 / DESIGNED FOR BUILDERS
    COMPILED LANGUAGE & DATA ENGINE

    Familiar thinking.
    Native execution.

    Write clear, typed, Python-shaped code. Compile the work into a native program, with a data engine designed around memory-conscious execution.

    • Native execution without the Python interpreter
    • Explicit parallel kernels and typed data operations
    • Notebook, editor and local model integrations
    first_kernel.ipynb
    ILLUSTRATIVE OUTPUT42✓
    Target notebook interface · release in preparation
    03 / THE PROVING GROUND

    Let the workload
    make the case.

    Inspect recorded results, change the constraint, and see the trade-offs. The test conditions stay in view.

    RECORDED RESULTSNATIVE BASELINE / 0.12.0
    128 MB CONTAINER LIMIT

    Completion under constraint.

    Native peak RSS105 MiB
    Native query time81 s
    Rows in fixture100 million
    CorrectnessMatched completed runs

    The competing engines were faster when uncapped. These results describe this fixture and baseline only.

    REPRODUCIBILITY HARNESS · LAUNCH KIT
    2 workers · 1 trial per capped cellDownload evidence notes ↓

    This explorer displays recorded measurements. Changing a control does not run an engine or simulate a new benchmark.

    BUILT AROUND THE WORK

    Your constraints.
    Your use case.

    Start with a representative dataset and a clear definition of success.

    ↯UTILITIES & METER DATA

    Turn intervals
    into useful answers.

    Explore meter aggregation today. Shape VEE, missing-reading review and revenue-anomaly workflows through a design partnership.

    ⌘DATA & PLATFORM TEAMS

    Keep the pipeline.
    Rethink the footprint.

    Evaluate compiled kernels and native SQL against your actual container limits, query shapes and correctness requirements.

    ⌁INDUSTRY & THE EDGE

    Bring the analysis
    closer to the plant.

    Scope local telemetry rollups and on-premises analytics. Develop future anomaly intelligence around real industrial signals. Use VexlFlow to explore order-to-test-to-dispatch process paths.

    04 / DEPLOYMENT ECONOMICS

    Size the workload.
    Then size the bill.

    Explore how resources, runtime and licensing affect your monthly cost. Every assumption is yours to change.

    01 / YOUR ASSUMPTIONS

    Edit infrastructure unit rates +

    Illustrative rates, editable to match your provider quote. Both scenarios use the selected vCPU count.

    Equal runtimes are the starting assumption. Validate the proposed resource and runtime values on your workload.

    02 / MONTHLY PLANNING ESTIMATE
    Current infrastructure$240
    Proposed total, with software$194
    Estimated monthly savings$46
    19.2%
    Proposed infrastructure
    $165
    Proposed software
    $29
    Annual cost difference
    $552 lower

    Planning model, not a cloud quote or benchmark. Excludes tax, storage, network, support and migration costs. Usage pricing adds the software fee to infrastructure. No live AWS or Azure pricing is fetched.

    05 / YOUR COMMERCIAL STARTING POINT

    Pick a plan. Prove the fit.

    Clear entry points for developers, analytical SQL, process mining and private intelligence.

    LAUNCH PLANNING PRICES · USD

    AWS / Azure Marketplace procurement and private offers.
    Listings are planned; request a scoped commercial evaluation.

    Launch planning prices in USD. Annual budgets show 12 months at the monthly rate, with no discount applied. Corporate scope, product availability, entitlements and billing terms are confirmed by agreement. The current SQL preview uses up to two execution lanes.

    THE QUESTIONS THAT MATTER

    A clear path
    to evaluation.

    Understand the fit before you commit.

    Where should I start: VexlPy, VexlSQL, VexlFlow or VexlSmart?+

    Choose VexlPy for typed compiled programs and native data operations. Choose VexlSQL for analytical queries and persistent tables. Choose VexlFlow to discover process paths, variants and handoffs in event logs. VexlSmart is the planned intelligence layer for SQL assistance, meter VEE and industrial anomalies.

    Is VexlFlow an alternative to Celonis?+

    VexlFlow targets native, local process mining. The current developer preview analyses case-centric CSV event logs and provides process maps, variants, elapsed handoffs and case timelines. Enterprise connectors, conformance checking, object-centric mining and multi-user governance are planned. Start with a scoped evaluation; this preview does not claim complete Celonis feature parity, automatic migration or a measured speed or cost advantage.

    Which client processes can I explore in VexlFlow?+

    Try fictional cross-border payments, home mortgages, bank customer service, procure-to-pay, order-to-cash, meter-to-cash, transformer production and mine-to-port scenarios. Each includes a prospective buyer, an investigation question and source-data requirements. Download its fictional CSV to reproduce the analysis with the native CLI, or explore its saved report here. The examples contain no actual customer transactions and establish no client performance or savings.

    Is VexlPy a drop-in replacement for Python and pandas?+

    VexlPy is a Python-shaped compiled language with typed native execution. It does not imply compatibility with every Python package or pandas API. Evaluate the supported syntax and integrations against your application.

    Does the proving ground execute a benchmark in my browser?+

    It displays recorded engineering results. No engine, dataset or AI model runs in the page. Each workload identifies its version, memory conditions and sampling method. Download the notes for the limits of each comparison.

    What is available in the SQL preview?+

    The native 0.4 preview includes analytical SQL, persistent tables, atomic write batches, compressed column storage with selective reads and min/max pruning, streaming HTTP and a limited PostgreSQL protocol adapter. Linux x86_64 was validated. The adapter uses loopback and autocommit; broader PostgreSQL compatibility, hardware encryption and vector search remain future work.

    Can I use Parquet, S3, Kafka or my SCADA historian?+

    These are integration opportunities to scope with your workload. Parquet and the advertised connector ecosystem are not verified shipping features in the reviewed previews. Current native SQL storage supports owned CSV, binary rows and optional compressed column blocks.

    How do private AI and air-gapped deployments work?+

    The SQL preview has an optional adapter for a configured model endpoint. It proposes SQL for human review. A complete air-gapped VexlSmart product, model quality, production VEE and anomaly workflows require further implementation and validation.

    How do checkout and enterprise proposals work?+

    Pro-4 and Pro-8 open a dedicated Paddle checkout path. A licence is issued only after the service verifies a completed payment. Checkout needs the merchant configuration before it can accept payments. Enterprise evaluations use a scoped brief; the $25,000 partnership has a draft proposal builder. Product release status is shown separately.

    VEXL SYSTEMS / DESIGN PARTNERSHIPS

    Bring your hardest
    data problem.

    For financial services, utilities, platform teams and industrial operators ready to scope a measurable evaluation.
    $25,000 for six months: one scoped data connector, five dedicated engineering days, and a shared validation plan.

    COMPILED COMPUTE · DATA SYSTEMS · PROCESS MINING · PRIVATE INTELLIGENCE
    LET’S DEFINE THE WORK

    Your next deployment
    starts with a brief.

    Define your workload and create a brief for a scoped product evaluation. Developer purchases use checkout; design partnerships have a separate proposal builder.

    This creates a text file in your browser. Your details are not sent or saved to a server, and no enquiry is submitted.
    VEXL SYSTEMS / SECURE CHECKOUT

    Your next build
    starts here.

    VexlPy Pro-4$29/ month

    One developer · up to 4 workers

    1. 01Complete payment in Paddle’s hosted checkout.
    2. 02A verified payment event activates your entitlement.
    3. 03Download your signed licence in the client portal.

    Monthly subscription in USD. Tax and final terms are shown by Paddle before payment. Sandbox purchases produce test licences. Product packaging and release status are separate from billing.
    VEXL SYSTEMS / CLIENT PORTAL

    Your access.
    Independently verified.

    Recover a purchase with the private receipt saved in this browser tab, or paste a recovery receipt you downloaded at checkout. Your receipt authorises access to your licence.

    No purchase selected.

    Browser checkout events do not issue licences. This portal waits for the verified payment webhook. A saved tab receipt expires after 30 minutes. Download your private recovery receipt to keep it; clearing or closing the tab removes the in-memory copy.
    DESIGN PARTNERSHIP / $25,000 USD

    One hard problem.
    A shared engineering plan.

    Build a reviewable, one-page engagement proposal. Six months, one scoped data connector and five dedicated engineering days.

    Draft for review. Payment schedule, acceptance criteria, IP, confidentiality, support and liability require a signed agreement. Creating a proposal does not submit it or commit either party.