JARQS AI LAB

Personal Operations Lab

Human-led systems for AI-assisted operations.

Monitoring, automation, data aggregation, and process intelligence shaped by practical IT experience and careful AI collaboration.

self-hosted mindset process-first monitored systems human accountable
Explore the lab Contact
About

Built by Jarek Koscielecki.

I build small, focused operational tools with AI as a design and implementation partner. My background is in IT systems, service operations, process architecture, and practical automation.

This lab documents a private ecosystem of experiments around monitoring, decision support, data aggregation, and everyday workflows. Public concepts, private operations.

Principles

Operational rules for useful AI systems.

The lab favors tools that are explainable, monitored, bounded, and shaped by real use.

01

Small tools over monoliths

Focused systems are easier to understand, audit, and evolve.

02

Monitoring from day one

Every useful tool needs health, alerts, and operational visibility.

03

Human control over automation

AI can accelerate work, but decisions stay accountable.

04

Private by default

Personal infrastructure should expose only what is intentional.

Lab Modules

A private ecosystem, described publicly.

These are conceptual descriptions only. Private deployments are not linked from this site.

00

Harem

Ecosystem console for apps, status, backlog, and operational overview.

01

Klaudia

Multi-model AI interface for structured conversations and experiments.

02

Szynka

Rail delay monitor with notifications and operational safeguards.

03

Agnes

News intelligence and source aggregation pipeline.

04

Medea

Appointment monitoring workflow for availability checks and alerts.

06

Penny

Price intelligence assistant for tracking offers and market signals.

Method

A loop for building practical tools.

  1. Observe

    Find a repetitive, fragile, or time-sensitive workflow.

  2. Design

    Model the process, data sources, risks, and failure modes.

  3. Build

    Create a small tool with clear boundaries and simple deployment.

  4. Audit

    Review quality, safety, cost, and operational risk before expanding.

  5. Monitor

    Add health checks, alerts, and feedback loops.

  6. Iterate

    Improve based on real use, not imagined scale.

Contact

AI-assisted operations, built carefully.

Want to talk about practical AI workflows, monitoring, or process automation?

[email protected]