Now booking one-workflow pilots. Our standard: a first agent live within 30 days of design sign-off.One-workflow pilots now booking. How it works
Book a Discovery Call

For ecommerce and customer-service leaders

Let AI answer “Where’s my order?”

AI agents that answer order and delivery questions, handle returns and damage claims, keep product pages accurate and audit ad spend, connected to your store, ERP and helpdesk. Your team handles the exceptions.

Book a Discovery Call See what we automate

45 minutes · a written plan, whether or not we work together

60%+ fewer tickets at one retailerWhatsApp, email & chatShopify, ERP & 3PL
Order-status agent · example run

Where margin leaks

Support queues and returns eat the margin.

Especially for home goods, furniture and other bulky or made-to-order products, and for any brand where “where is my order?” dominates the queue.

32

support tickets per 100 orders for Home & Garden brands, against a 25 median across verticals1

22.7%

online return rate for furniture, against 19.3% across all categories2

$55–108

cost to process one large-item return, often 50–100% of its gross margin3

$1,127

average Google Ads spend wasted per account per month4

Sources: 1 Gorgias Ecom Lab via Macha (opens in a new tab) · 2 NRF & Happy Returns via eightx (opens in a new tab) · 3 Optoro via eightx (opens in a new tab) · 4 WordStream via PPC Land (opens in a new tab)

What we automate

Five workflows, starting where the volume is.

We start with the work that is high-volume and rules-based, where every hour saved shows up in your support costs or your margin.

  1. 01

    Order and delivery status

    Answers “where is my order?” on WhatsApp, email and chat from live store, ERP and carrier data, including made-to-order production status and white-glove delivery windows.

    Fewer tickets, faster replies at peak
  2. 02

    Returns and damage claims

    Collects photos and order details, checks your returns policy, books collections and routes refunds above your threshold to a person.

    Claims logged in minutes, not days
  3. 03

    Catalog and product-page accuracy

    Checks every product page for missing dimensions, materials, care details and inconsistent specs, and drafts fixes for approval.

    Fewer “not as described” returns
  4. 04

    Ad spend audits

    Watches Google and Meta accounts daily for wasted search terms, budget pacing problems and falling ROAS, with plain-English alerts.

    Spend moved to what converts
  5. 05

    Customer sentiment

    Reads reviews, tickets and chats to flag product, delivery and service issues before they become a pattern.

    Problems caught early
Connects to

Logos show integration compatibility, not partnership.

Before and after

What changes in a peak week.

  1. Order status

    Today: Support staff copy tracking numbers from three systems into replies, and queues double in peak season.

    With agents: Routine status questions are answered instantly from live data. Your team handles only real exceptions.

  2. Returns

    Today: Claims arrive as emails with missing photos, and each one takes several back-and-forths.

    With agents: Every claim arrives complete, policy-checked and ready for a one-click decision.

  3. Product data

    Today: Missing dimensions and mismatched materials are found when a customer returns the item.

    With agents: Gaps and inconsistencies are flagged and fixed before the product goes live.

  4. Ad spend

    Today: Wasted spend is found at the monthly review, after it has gone.

    With agents: Problems are flagged the same day, with the fix suggested.

Results

Already live in production.

From retail and adjacent content work, measured in the client’s own terms.

  • Automation60%+Delivery-support tickets cut by more than 60% with a WhatsApp order-status agent, with 10,000+ queries automated every month.National food & fashion retailer · 900+ stores · South Africa
  • Automation70% fasterA briefing agent turns scattered client material into structured, publishable campaign briefs. Brief creation became 70% faster.Creator-marketing SaaS · Australia

They stuck to the timetable, completing each milestone on time, with clear, measurable results throughout.

Client review, Real Estate · AI consulting engagement
See all resultsResults as reported from our engagements. Ask us for a reference call.

How engagements work

Start with one workflow. Scale what works.

Order status is the natural first workflow: high volume, clear rules and results you can measure in weeks.

  1. Step 01

    Discovery call

    45 minutes on your highest-volume workflows. You get a written plan: what to automate first, in what order, and what it should return.

  2. Step 02

    Pilot on one workflow

    We build and run one agent on your own accounts, against real cases, with your people approving its output. You see the numbers before committing further. Duration and commercials are agreed on the discovery call.

  3. Step 03

    Scale and tune

    Proven agents roll out to more teams and workflows on a monthly retainer, tuned as your volumes, rules and systems change.

Questions

Questions retail teams ask.

Do we have to replace our helpdesk?

No. Agents work alongside Gorgias, Zendesk, Freshdesk or your own inbox, and connect to Shopify, WooCommerce, Magento, your ERP and your carriers through their APIs. We confirm exactly what connects on the discovery call.

Can it handle bulky, made-to-order or white-glove delivery?

Yes, and that is where off-the-shelf tracking tools fall short. We connect to your production schedule, 3PL and delivery partners, so customers get real answers about build and delivery dates, not just a parcel tracking link.

Will customers know they are talking to an agent?

We recommend telling them, and some jurisdictions require it. The agent keeps your brand voice, and anything sensitive, such as refunds, complaints or high-value orders, goes to a person with the full context attached.

Can we go live before peak season?

Usually, if we start now. Our standard is a first agent live within 30 days of design sign-off. We run it on a share of conversations first, tune it to your policies and tone, then scale it up before the peak arrives.

How is our data protected?

Agents run with scoped, revocable permissions, every action is logged and changes to records can be rolled back. Where your policies require it, we deploy into your own cloud, and model providers are configured so your data is not used to train their models. Our security management is certified to ISO 27001.

What does it cost?

We don’t publish a price list for this work. It depends on the workflows, volumes and systems involved. Pilot duration and commercials are agreed on the discovery call, based on your scope and requirements, and you get a written proposal with a fixed scope and price before any build starts.

Be ready before your next peak.

Book a 45-minute discovery call. We’ll look at your ticket mix, returns and ad accounts, and send you a written plan for the first agents to build.

Book a Discovery Call

Not ready for a call? Get your AI Maturity Score in 2 minutes

Book a Discovery Call