Inside Brarista's Onboarding: How We Build Your Personalised AI Bra Fitting Experience from the Ground Up

Everywhere you look, retail brands are adding AI to their customer experience. From fit calculators to conversational AI chatbots, there's no shortage of tools promising to solve the sizing problem online. It's hard to avoid the conversation, and harder still to know where to start, as teams barely have any bandwidth to take on a whole tech project.

And we get you - Brarista was built from the ground up whilst consulting with the lingerie industry to identify exactly these pain points. Whether you're switching from another tool, starting fresh, or building on top of what you already have, we've designed our onboarding to be as stress-free as possible.

That said, as a conversational AI tool built exclusively for intimate apparel, we know how complex this category really is - and we believe getting things right starts at the product level. So while we keep the process simple on your end, we do meticulous work behind the scenes on your product catalogue, size charts, and brand-specific fit logic before anything goes live.

Whether you're launching our Full Chatbot, Mini Chatbot, or Sidebar Fitting Quizzes, the experience your shoppers get is only as good as the data and logic behind it. We review inconsistencies in your sizing, stress-test your product data, and in many cases, brands have updated their own size charts off the back of what we found.

We spoke to the people behind our onboarding process - our QA specialists, visual designer, back-end and front-end engineers, and our founder Bella, who comes from a bra fitting background herself - to walk you through what this looks like from the inside.

Who is in charge of my onboarding, and after?

A team photo or individual headshots with name/role labels

The people who set you up during onboarding are the same specialists who continue to support you after launch, and that matters more than it may sound.

Lingerie sizing is not something you can hand to a generalist. The nuances of sister sizing, international size conversions, cross-brand fit logic, and breast shape assessment require people who have spent time understanding the industry. That is why fitting knowledge runs through the whole process - from the baseline of specialist lingerie knowledge built over years with professional fitters, to the brand-specific layer we build during onboarding, where our team consults directly on your size chart, product catalogue, and fit logic. We even use fit models to physically verify recommendations before anything goes live. As Zeynep, our QA and client success lead, puts it: "I always keep the consumer and lingerie wearer point of view in mind. No interaction with the bot is meaningless or not important."

Bella, our founder and CEO, is one of the fitting experts involved from the start, bringing 6+ years of professional bra fitting expertise to the sizing decisions that underpin the product logic. Alongside her, our QA specialists, visual designer, front-end and back-end engineers each play a distinct role in making sure the tool is right before it reaches your shoppers.

What does our team actually need to do during onboarding?

Brarista's three-step onboarding process: AI customisation, demo and integration, and launch

With fewer than 20 hours required from your team, very little.

Setting up a generic AI agent typically means connecting your help centre, writing brand guidance, uploading documents, defining rules, and then refining the AI over time as it learns - often with your team doing much of that ongoing work.

With Brarista, it's different. As Bella explains: "We've designed onboarding to be as hands-off as possible because we know you're running a business, not managing tech projects. Give us access to your Shopify - or whatever platform you're on - and fill out our onboarding form. That's the heavy lifting done."

Beyond that, your team's role is to test the product and confirm that the answers align with your sizing philosophy - because that alignment needs to come from you. All the technical setup: catalogue integration, AI training on your products, tailoring, and deployment is handled by us.

We get brands live in as little as two weeks

Here's what typically happens on our side during those two weeks:

Step 1: AI customisation

This is where you share your product catalogue and customisation requirements with us - but don't let the word "form" put you off. This is where the in-depth fitting work begins, and it's the most involved part of the whole process - on our end, not yours.

Size mapping is where the team spends the most time. As Zeynep describes it: "We make sure that users get recommended the correct size by measuring the products, trying them on true-to-size models, comparing them to other products that have set the industry standard - and sometimes even trying them on ourselves."

Our team at work - every product in your catalogue gets this level of attention before go-live

On the back-end, our engineering lead works through your product catalogue to make sure every item is understood thoroughly - tagged with the fit and sizing logic that allows the tool to make appropriate recommendations. "We've built our system so that sizing is always accurate and consistent without hallucination - the AI handles the conversation, but all size calculations follow strict brand-specific rules we've built ourselves. It cannot recommend a size that doesn't exist in your catalogue," our back-end lead explains. For a deeper look at how that works, this post covers it in full.

Most brands arrive with size calculators and measurement charts already live on their site. If those don't align with what the bot recommends, shoppers get conflicting information - which is worse than no guidance at all. "We make sure that the client understands our sizing philosophy - and make the necessary changes to their own sizing charts if needed," Zeynep adds.

Latched's sizing overhaul, developed in collaboration with our team during onboarding

It is also where we can add real value for brands that are still developing their sizing, or struggling with returns due to sizing inconsistencies. Somewhere Never updated their sizing after reviewing the detailed report we produced during onboarding, as they found their original system was creating unnecessary confusion for shoppers. For Latched, our maternity activewear partner, we went a step further and collaborated to overhaul their product sizing charts entirely before going live.

Step 2: Design and integration

While sizing logic is being built and tested, our visual designer and front-end engineer are working in parallel on how Brarista looks and feels on your site. Colours, fonts, tone of voice, the chatbot name, the questions you want to surface - these are some of the many things you can get tailored. Because we adapt so thoroughly across every client's site, you won't be able to tell where Brarista has been integrated, unless you look closely enough.

Brarista chatbot integrated across multiple lingerie retailer websites, each reflecting the brand's own visual identity

Our front-end engineer Khoa starts by reviewing your platform and existing theme - fonts, colours, overall styling - so the integration feels native from day one. "Every client has a slightly different setup," he explains. "We manually review each client's codebase to position the chatbot components in the right place within their theme. If the platform has been onboarded before, setup is usually completed within a day." We currently work across Shopify, BigCommerce, and Wix, and have an established testing process on each.

Our visual designer Anna adds: "We've worked with clients that are early on in their branding phase and don't have a strict design system in place - which is not a problem, as we align the assets to their existing site. These front-end aspects can easily be changed upon the client's request, so even after onboarding, if there's a seasonal colour update or a brand refresh, we can do a quick swap."

The product exists as a single line of code in your website. Testing is done in a separate test theme within your platform, so your live site is never disrupted before thorough verification. The only access we need from you is to your host platform - and only for the duration of the integration.

Step 3: QA, testing, and launch

Before anything goes live, Zeynep and the QA team put the tool through its paces across three categories - FitAI, StyleAI, and CareAI.

"We test the product on different devices, browsers, and internet speeds. We test different sizes - rare sizes, out-of-range sizes, alternative sizes. We test misspelled queries for accuracy. And sometimes we even order products using the bot to live the real shopping experience and try them for fit," Zeynep describes.

This is also the stage where your team gets most involved - not to manage the technical side, but to sense-check the output against your own knowledge of your products and sizing. If something doesn't match your sizing philosophy or how you'd expect a particular query to be answered, that's a conversation we have before go-live and continue to have after. We handle all the adjustments throughout.

What happens after we go live?

See more of our clients' experiences with onboarding on our wall of love

Going live is not the end of the process - it's the start of the partnership.

No testing process can anticipate every real-world scenario. Once a brand goes live, we keep monitoring conversation logs closely - because some edge cases only surface when real shoppers start asking real questions. When they do, we move quickly to make necessary adjustments. When a new product range is introduced or there is a seasonal rebrand, we make sure the fit logic and knowledge base are updated on an ongoing basis, so the tool gets sharper the longer it's live.

Your team gets visibility into all of this through the admin dashboard - built around the KPIs we found matter most to lingerie retailers specifically, including bra wearing habits data that generic e-commerce tools do not track.

As Bella puts it: "We'll launch and monitor the first three months on a daily basis to make sure everything works. And if we don't go live due to our technical limitations - it's a full refund on the onboarding fees."

What if our brand has unconventional sizes?

That's precisely where we shine. If you're offering extended sizing, custom measurements, or sizing that doesn't follow standard conventions, you're exactly the kind of brand we want to work with.

Conventional sizing is actually the minority in lingerie. Most brands have quirks - some run small in the band, others cut cups differently, extended size specialists use completely different grading systems. During onboarding, our team learns your sizing structure: how your bands fit, whether your cups run true or generous, what makes a particular collection different. The big generic tools can't handle sizing outside narrow parameters. We can, because we're built by people who understand that unconventional sizing isn't a problem to work around - it's just how lingerie retail works.

What if we're a multi-brand or multi-sizing retailer?

This is one of our most distinctive capabilities. A 34C in one brand doesn't fit the same as a 34C in another - your customers know this from experience, and most fitting tools pretend otherwise. That's why returns stay high.

Brarista handles product-specific size mapping across your entire catalogue. When a customer asks for a recommendation, we tell them exactly what size they need in each specific product. If they're a 34C in one brand but need a 32D in another, we know that and surface it directly on the product card. Your customers aren't just asking "what's my size?" - they're asking "what's my size in this specific bra?" We're built to answer that question accurately across multiple brands, which is what keeps returns low and shoppers confident.

Ready to see what this looks like for your brand?

Onboarding starts with a conversation. Book a demo and we’ll walk you through what the process would look like for your specific catalogue, platform, and sizing set-up.

About Brarista

Brarista is the world's first AI-powered bra fitting chatbot platform designed specifically for lingerie retail. By merging advanced AI technology with professional bra fitting expertise, Brarista revolutionises the online shopping experience and tackles the long-standing challenge of incorrect bra sizing, which affects 90% of consumers and drives dissatisfaction and sales drop-off.

Unlike generic chatbots or simple measuring tools, Brarista offers deep lingerie specialisation — combining professionally graded fitting quizzes, deep learning algorithms, and large language models to democratise expert fitting services at scale. It delivers round-the-clock, personalised, brand-agnostic sizing recommendations through B2B2C white-label software embedded directly within retailers' websites, free to end-users.