Skip to content
Zenet Labs

Automation · Integrations · AI agents

Stop doing by hand what your tools can do on their own.

We connect the tools your company already uses (ERP, online shop, CRM, WhatsApp, spreadsheets) and put AI agents on the repetitive work, so your team stops moving data by hand.

For companies of 10 to 100 people

Illustrative example of an agent run log: shop orders turned into invoices, supplier invoices matched, customer questions answered and a weekly report in preparation.

01 — Sound familiar?

Three jobs your team shouldn't be doing by hand

  • Orders retyped by hand

    Every order from the shop or a marketplace is copied into the ERP or a spreadsheet. One typo and the wrong item ships.

    What we do

    We connect the shop to your ERP and invoicing, so each order arrives once, complete, with stock updated.

  • Invoices typed in

    Supplier invoices arrive as PDFs and someone keys them into the accounting system, line by line.

    What we do

    An agent reads each invoice, matches it to the order and leaves only the exceptions for a person.

  • Customers served one by one

    The same questions all day on WhatsApp and email: where is my order, is it in stock, how much is it?

    What we do

    An agent answers the routine questions with data from your systems and hands the rest to your team, with context.

03 — How we work

Start small, prove it, then scale

  1. Diagnose

    In one to two weeks we map your tools and manual work and hand you a costed plan.

  2. Pilot

    We automate one process end to end, in production, with your team, before you commit to more.

  3. Scale

    We extend what worked to more processes and tools, with monitoring and approval steps in place.

  4. Maintain

    We monitor, fix and improve every month, so automations keep working as your tools change.

04 — Cases

Tested on our own companies first

Before we build an automation for a client, we run it in a business we own.

  • Travel · Small-group tours to Asia

    Viatsy

    Viatsy runs small-group trips to Japan, China, Korea, Taiwan and Hong Kong. Every trip means inquiries on several channels, a tailored proposal and many supplier bookings.

    What runs on its own

    • A drafted first reply to every inquiry, which a person reviews and sends
    • Proposals drafted from the client's brief and the trip catalogue
    • Supplier confirmations read into the booking, and the client's travel pack assembled from them
    • Payment requests and reminders sent on schedule
  • SaaS · Software for travel agencies

    Taplany

    Taplany is the platform we built from Viatsy's automations so other travel agencies can use them: website, inquiries, proposals, bookings, payments and WhatsApp in one place.

    What runs on its own

    • Replies to inquiries drafted in the agency's tone, for a person to approve
    • Itineraries and supplier confirmations pasted as text turned into structured bookings
    • WhatsApp, email and payment links in the same timeline as each booking
    • Website content drafted and fact-checked before it is published

05 — Founder

You work with the person who builds it

Albert Sabaté

Founder

15+ years building and leading engineering teams, including leading engineering at Splyt as its global mobility network scaled across several countries, and core contributions to Brisa, an open-source web framework. Today he runs Viatsy and Taplany with agents doing the repetitive work.

More about Albert →

06 — FAQ

What people ask us most

What do you do, and what don't you do?

We connect the tools your company already uses, automate the repetitive work between them and put AI agents into production for tasks such as triage, replies and reading documents. We also build custom web apps, train teams and maintain what we build. We don't resell software licences, run paid ads or replace your accountant or IT support.

Which companies is this for?

Companies of about 10 to 100 people that run on several tools that don't talk to each other, such as an accounting system, an online shop, a CRM, WhatsApp and spreadsheets, and lose hours copying data between them. A smaller team can start with a single integration.

Which tools do you work with?

Most often Xero, Shopify, WooCommerce, HubSpot, Google Workspace, Microsoft 365, WhatsApp Business and spreadsheets. If a tool has an API or a reliable export, we can usually connect it. The diagnostic confirms what is possible before you commit to anything.

Do we have to change the tools we use?

No. We work with what you already have and suggest a change only when a tool is the bottleneck, for example when it can't connect to anything else. Any change we suggest comes with its cost, so you decide.

How long does a project take?

The diagnostic takes one to two weeks. Its plan sets a timeline for each phase, and we scope the first delivery so it goes live in weeks, not months. Larger agents and custom apps are split into phases, each with its own result.

How much does it cost?

The diagnostic costs S$3,000, credited to the project if you continue within 60 days. An integration or automation costs from S$8,000, an AI agent in production S$15,000–35,000 and a custom web or app from S$12,000. Prices exclude GST. Training, maintenance and growth are on the pricing page.

What does maintenance include?

Monitoring and error alerts, fixes, updates when one of your tools changes, and a monthly review of what to improve. The three monthly plans differ in how many automations and agents they cover and how much improvement work they include.

Who owns the code and the data?

You do. We build on your accounts wherever we can, the code we write for you is yours once the project is paid, and your data stays in your systems. If we stop working together, everything keeps running and is documented.

Where is our data kept, and how do you handle privacy?

We agree where data lives before we build, and we list every provider that will touch it. We work under Singapore's PDPA and, for clients in Europe, the GDPR, with a data processing agreement. We use AI providers that don't train on your data, and we tell you early if a use case is regulated under the EU AI Act, such as screening job candidates.

What happens if an agent makes a mistake?

We plan for it. Agents work within limits you agree, actions with consequences, such as payments, refunds or messages to key customers, wait for a person's approval, and every action is logged. When something goes wrong, monitoring shows it, we fix the cause and add the case to the agent's tests.

Which AI models do you use?

Whichever does the task reliably at the lowest cost and meets your privacy needs: models from providers such as OpenAI, Anthropic or Google, or open models hosted in a region you choose. We test them on your real cases before choosing.

Do you train our team?

Yes. Workshops are hands-on and use your own processes, not generic examples. Monthly mentoring keeps the team going afterwards: we review what they build, help when they get stuck and agree simple rules for using AI at work.

How do we know it's working?

In the diagnostic we measure how long each process takes today. After go-live we compare the same numbers, such as hours spent, errors and response times. If an automation doesn't pay for itself, we tell you.

Do you work with companies outside Singapore?

Yes. We work remotely with companies in Singapore, Spain and elsewhere, in English or Spanish. Prices on this site are in Singapore dollars; the Spanish version shows prices for Spain in euros.

How do we start?

Book a diagnostic with the form on this page or write to hello@zenetlabs.com. We reply personally, agree the scope on a short call and, one to two weeks later, you have a costed plan. If you go ahead within 60 days, the diagnostic fee is credited to the project.

07 — Contact

Book your automation diagnostic.

Tell us which tools you use and what your team still does by hand. We reply personally and agree the next step with you.

Prefer email? Write to hello@zenetlabs.com

A human replies, personally.