Skip to content

Home

Local AI-Native RPA in Python

Zero-license unattended automation, human-readable Markdown flows, and privacy-first local intelligence. Automate web, Excel, and desktop workflows without per-bot fees.

Get Started → GitHub pip install kinenix

Why Kinenix?

  • Zero-License Unattended RPA

    Run unattended bots across 1 or 1,000 servers without per-bot licensing fees or hidden orchestration taxes. True open-source automation designed for production scale.

  • Local AI-Native (Privacy-First)

    Powered by local Small Language Models (SLMs) via Ollama and local CPUs. Process sensitive financial, HR, and customer documents with zero data leaks to third-party cloud APIs.

  • Zero-License Office Automation

    High-speed Excel, CSV, and tabular data automation powered by native Python engines (openpyxl, pandas). No Microsoft Office installation or desktop licenses required.

  • Deterministic + Agentic

    Eliminate unpredictable hallucinations. Workflows are declared in structured, readable flow.md files that execute deterministically, with AI invoked only where intelligent extraction or self-healing is needed.

Human-Readable Flows (flow.md)

In Kinenix, every workflow is a simple, version-controllable Markdown document. It can be read, written, and maintained by developers, business analysts, and AI assistants alike:

# Flow: Download Daily Exchange Rates

A daily automated flow to scrape foreign exchange rates and update our ERP Excel sheet.

## Steps

1. **Open Exchange Rates Portal**
   - Action: `web.open`
   - URL: `https://example.com/fx-rates`
   - Headless: `true`

2. **Extract Latest USD/THB Rate**
   - Action: `web.get_text`
   - Selector: `#usd-rate`
   - Output: `usd_rate`

3. **Update Master Excel File**
   - Action: `excel.set_cell`
   - File: `data/accounting.xlsx`
   - Sheet: `Rates`
   - Cell: `B2`
   - Value: `${usd_rate}`

4. **Send Notification**
   - Action: `email.send`
   - To: `finance-team@company.com`
   - Subject: `Daily Rates Updated: USD = ${usd_rate}`

Comparison: Kinenix vs. Alternatives

Feature Kinenix Enterprise RPA (UiPath / Power Automate) Raw Python Scripts
Licensing Cost Free & Open Source (Zero per-bot fee) \$5,000 - \$15,000+ per bot / year Free
Flow Format Readable Markdown (flow.md) Proprietary XML / JSON / Low-Code UI Raw Code (.py)
AI Integration Native Local SLM (Ollama) / Central LLM Cloud-only add-on (extra cost) Manual SDK integration
Data Privacy 100% Local execution (PDPA/GDPR safe) Often requires cloud telemetry/APIs Local
Unattended Runner Built-in daemon with schedules & watchers Expensive Orchestrator add-on Requires systemd / cron / Celery setup
Central Telemetry Lightweight Self-Hosted Hub (kinenix hub) Heavy enterprise server Custom database & monitoring required

Modular Architecture

flowchart TD
    subgraph Developer & AI Workspace
        A["flow.md Definition"] --> B["kinenix run (Core CLI)"]
    end

    subgraph Production Unattended
        W["kinenix-worker (Schedule / File-watch)"] --> B
    end

    subgraph Actions Engine
        B --> C1["web.* (Playwright)"]
        B --> C2["excel.* (openpyxl/pandas)"]
        B --> C3["ai.* (Ollama / Local SLM)"]
        B --> C4["file.* / logic.*"]
    end

    subgraph Central Telemetry
        B -. Push Telemetry .-> H["kinenix-hub (FastAPI + SQLite/Postgres)"]
        W -. Heartbeats .-> H
        H --> D["Audit Log & Dashboard (www.kinenix.com)"]
    end

Quick Start in 60 Seconds

Install the core framework and create your first automation:

# 1. Install Kinenix
pip install kinenix

# 2. Create a new task project
kinenix init "Process Daily Invoices"

# 3. Validate your flow
kinenix validate .

# 4. Execute the flow
kinenix run .

# Install with hub and worker daemon support
pip install "kinenix[hub,worker]"

# Start the central telemetry hub
kinenix hub

# Connect this machine's runs to the Hub
kinenix connect http://localhost:8080