Tatiana Beauty is a booking and operations system built for a real makeup artist and educator in Chișinău, Moldova.

Most customer inquiries start on Instagram or WhatsApp with questions like “Are you free Saturday?”, “How much is makeup?” or “Can I book makeup and hair?”

What looks like a simple question creates repetitive work: checking the calendar, proposing times, waiting for a reply, confirming the appointment, sending reminders, handling rescheduling and trying to refill cancelled slots.

Tatiana Beauty moves that coordination into one workflow.

Customers see real availability and request an appointment themselves. Tatiana keeps control with final booking approval, while confirmations, reminders and customer self-service handle the routine follow-up.

When a time becomes available, the waitlist can automatically find the oldest compatible customer, reserve the slot temporarily and offer it to them. If accepted, it converts safely into a normal booking request.

The system also coordinates combined Makeup + Hair appointments across Tatiana and Diana’s independent calendars.

For Tatiana, the private studio brings together appointments, manual Instagram/WhatsApp bookings, working hours, blocked time, customers, waitlist activity and operational impact.

The product principle is simple:

Tatiana manages the exceptions, not every step.😍

🚀 The pace of technological change is no longer linear, it's exponential.

From automation to Agentic AI, we're witnessing a transformative shift across industries — reshaping how we work, hire, and scale.

As part of this shift, I recently built an AI-powered recruitment screening workflow using n8n — one that could meaningfully support the Talent Acquisition function, not replace the judgment it depends on.

❓ PROBLEM Most companies still handle recruitment manually — screening CVs one by one, replying to applicants, checking for duplicates, and manually tracking which positions are still open. It's repetitive, slow, and error-prone — and it eats hours of HR's time every week.

💡 SOLUTION I built an AI-powered recruitment automation that: ✅ Automatically reads incoming applications from Gmail ✅ Matches and validates the position applied for against open roles — even with typos or inconsistent naming ✅ Extracts and evaluates each CV against HR-defined criteria using AI ✅ Detects duplicate applications and responds accordingly ✅ Informs applicants when a position is closed, and shares current openings ✅ Routes candidates into three tiers — Recommended, Consider, Reject ✅ Logs every application and outcome for full traceability ⚡️ RESULT (based on real testing across 8 scenarios) ✅ Processes a full application — reading the email, matching the position, extracting the CV, running the AI evaluation, and replying — in an average of ~25 seconds ✅ Early rejections like duplicates and invalid applications resolve in under 15 seconds ✅ Even non-standard CVs (image-based/creative formats) still get processed automatically, falling back to a secondary extraction method — takes ~40 seconds, still well under a minute ✅ Borderline ("Consider") candidates reach HR's review queue in roughly the same ~25-30 seconds — the only wait left is a human decision, not the system ✅ Every application, across all 8 tested scenarios, is logged and traceable — nothing gets lost or forgotten ✅ Runs continuously with no dependency on someone being available to trigger it

🧠 WHY IT'S DIFFERENT This isn't a "trust the AI blindly" system. AI can misread or hallucinate — so every ambiguous case gets flagged and routed to a human instead of being silently auto-rejected. The goal is removing repetitive work, not removing human judgment where it actually matters.

🔧 WHAT'S NEXT I'm continuing to refine the evaluation logic based on real test cases, and exploring how to extend this into other parts of the hiring pipeline.

💡 AI isn't just a tool that speeds up existing processes — used well, it becomes part of how a workflow makes decisions. The real work is designing where AI should lead, and where a human still needs to.

🤝 If you're exploring how n8n or AI-powered workflows could fit into your hiring process (or anywhere else in your ops), feel free to drop me a message — happy to connect and compare notes.

Check it out