Case studies

Real projects.
Real automation.

Every system we build solves a real problem for a real business. Here's what that looks like in practice.

Pazolini

Fashion & RetailItaly

The problem

Pazolini sells across multiple marketplaces and receives a high volume of customer inquiries daily. Customers ask about sizing, availability, shipping, returns, and order status across email and WhatsApp. The support team was overwhelmed, response times were slipping, and repetitive questions were eating up hours that could go toward actual problem-solving.

What we built

We built a fully automated customer support bot that handles inquiries across email and WhatsApp. The bot is deeply briefed on Pazolini's product catalogue, sizing guides, shipping policies, and return procedures. It processes incoming requests, answers product-specific questions accurately, and only escalates edge cases to the human team. One system, multiple channels, always on.

Stack:Claude APIWhatsApp Business APIEmail integrationn8nKnowledge base

Support response time dropped from hours to seconds. The team now only handles cases that actually need a human.

PZLN

B2B Marketing AgencyLuxembourg

The problem

PZLN is a B2B marketing agency that was struggling to find new clients. Their outreach was inconsistent, follow-ups were falling through the cracks, and there was no structured email marketing in place. They knew their ideal client profile but had no system to find and reach them at scale.

What we built

We built an automated lead generation pipeline. The system researches prospects that match PZLN's ideal client profile, qualifies them against specific criteria, and sends personalised outreach on autopilot. Follow-up sequences run automatically across multiple touchpoints, so no lead goes cold. PZLN reviews and approves messaging before it goes out.

Stack:Apollo.ioClaude APIn8nEmail sequencesCRM integration

Consistent pipeline of qualified prospects. Outreach runs daily without anyone pressing send.

PrimeKids

KindergartenLuxembourg

The problem

PrimeKids manages 50+ monthly invoices across multiple kindergarten locations. Every month, the finance team spent hours manually matching invoice line items against bank statement entries in Google Sheets, cross-referencing amounts, dates, and parent names to confirm which payments had been received and which were outstanding.

What we built

We built an AI agent that reads uploaded invoices and bank statements, automatically matches payments by amount, date, and client reference, and flags discrepancies for human review. The matched results are written directly into the existing Google Sheets workflow. No migration needed.

Stack:Claude APIGoogle Sheets APIn8nPDF parsing

Hours of weekly manual reconciliation reduced to minutes. Discrepancies caught automatically.

Brococo

Meal PrepSpain

The problem

Brococo needed to send weekly menus to their client base every Monday and notify customers about special promotions and events. The process was entirely manual. Copying menu text, formatting messages, and sending them one by one through Telegram.

What we built

We deployed a Telegram bot that automatically sends the weekly menu on a set schedule. The team updates the menu in a simple interface, and the bot handles formatting, delivery, and broadcast to all subscribed clients. Promotions can be triggered on-demand with a single message.

Stack:Telegram Bot APIn8nSupabaseScheduled automation

Zero manual messaging. Clients receive menus on time, every week, without anyone pressing send.

Ready to automate?

Book a 20-minute call. We'll find the workflow that's costing you the most time.

Book a call