Services · Mobile app development · AI chatbots
Codeeo builds AI chatbots for London businesses that field a steady stream of customer questions: West End retailers, letting agents, clinics, hotels, venues and professional firms in the City. The chatbot sits on your website and on channels such as WhatsApp and Instagram, answers from content you have approved, and passes the conversation to a person when a question needs one. Customers get a reply in seconds instead of hours, and your team stops typing the same answer twenty times a day.
We have no London shopfront. Our team of 80+ works from Dubai, Riyadh and Karachi and runs discovery, demos and reviews with London clients by video call. You watch the bot answer genuine questions from your own inbox before it goes anywhere near a customer.
200+ companies servedFrom a first store to an ADX listed developer
What the chatbot handles
Before we write any code we read a sample of your real enquiries, with personal details removed. In London that sample tends to show a familiar pattern: a large share of questions repeat, many arrive late in the evening, and plenty come from visitors and newcomers who do not yet know how things work locally. A good chatbot handles that repeat layer well and knows exactly when to stop. The capabilities below are the ones London clients ask about most, though a first release normally picks four or five of them. Everything the bot says comes from material you sign off, so it does not invent opening times, policies or availability. Where a question touches money, health, a complaint or a legal matter, the bot hands over to a person with the conversation attached instead of attempting an answer it should not give.
We connect the chatbot to your FAQs, policy pages, product data and help articles, then restrict its answers to that material. When nothing in the content covers a question, the bot says so and offers a human route. Sources refresh on a schedule, so a change to your returns policy or opening times reaches the chatbot without anyone retraining it.
Customers get answers your team would stand behind.
A light chat window that loads without slowing your pages, carries your brand and works with keyboards and screen readers to WCAG 2.2. It respects your cookie banner, so any tracking inside the widget only runs after a visitor has consented, in line with PECR.
A chat window that fits your site and your compliance rules.
Many Londoners would rather send a message than fill in a form. We connect the same chatbot to the WhatsApp Business Platform, Instagram direct messages and Facebook Messenger, with one knowledge base behind every channel, so a customer hears the same answer wherever they ask.
One set of answers across every messaging channel.
London customers write in a great many languages. The chatbot detects the language of a message and replies in it, while your approved English content stays the source of truth. We agree the supported languages in writing and have each one checked by a fluent speaker before launch.
Visitors and residents get help in the language they wrote in.
When a customer asks for a person, sounds frustrated or raises something sensitive, the chatbot passes the conversation to your inbox or help desk with a short summary. Staff see what was already said, so the customer never has to repeat the whole story.
Difficult conversations reach a human with context.
For letting agents, venues and professional firms, the chatbot gathers the details your team needs, such as dates, party size, postcode or the type of matter, and sends a tidy record to your CRM or shared mailbox. It only requests data you have a lawful basis to collect.
Leads arrive complete instead of half described.
The bot can read your booking system and tell a customer which tables, rooms or appointment slots are free, then link them to confirm. It answers the availability question only; taking payments or changing bookings stays with your existing systems unless you choose otherwise later.
Fewer calls asking whether there is space tonight.
A dashboard shows the questions people ask most, where the bot handed over, which answers customers rated poorly and which pages the questions came from. That tells you what to fix on the website itself, not only inside the chatbot.
Real customer questions become a clear list of content fixes.
Retention periods, redaction of card numbers and other personal details typed into chat, and a short privacy notice inside the widget are configured to UK GDPR. Your data protection officer can review the settings and the list of processors before anything goes live.
A chatbot your data protection lead can sign off.
Who this is for
A chatbot is worth building when the same questions arrive in volume and the answers already exist somewhere in your business. It is the wrong tool when every enquiry needs judgement. These are the London situations where the case is clearest, and in each one the first release stays narrow so you can measure it before adding more. If your enquiries do not fit the pattern, we will say so during discovery rather than sell you something that sits unused.
Shops around Oxford Street, Covent Garden and the big shopping centres get questions about stock, click and collect, returns and opening hours, often from tourists comparing several stores in one afternoon. A chatbot connected to your product feed and store pages answers those in seconds, in the visitor's own language, and points them to the right branch. Staff on the shop floor stop fielding the same phone calls, and online shoppers get an answer before they drift off to a competitor.
Product and store data must be current for this to work.
Tenants and applicants ask about viewings, holding deposits, how to report a repair and which documents a landlord needs for right to rent checks. Much of that has a standard answer. The chatbot covers the standard part, logs repair reports with photos for your property team, and passes anything about a dispute or arrears to a named person. For agencies working across several boroughs, it can also tell an applicant which branch looks after which postcode.
Disputes and arrears always go to a person.
Guests ask about check in times, luggage storage, step free access, the nearest Tube station and whether the kitchen can cope with an allergy. A chatbot answers most of that from your own guest information at any hour, in several languages, and passes allergy and access questions to staff with the details already captured. For venues on the South Bank or in Shoreditch, event day questions about entry times and bag policies are the obvious first set to automate.
Allergy answers are confirmed by your team, never guessed.
Law firms, accountants, wealth managers and insurance brokers in the Square Mile and Canary Wharf want a chatbot that sorts enquiries without giving advice. We design it to explain your services, collect the facts of a new enquiry, pass names for conflict checks to your intake team and offer a call. Firms regulated by the FCA or the SRA get wording that stays clear of advice and financial promotions, plus conversation records that suit their compliance reviews.
The bot introduces your people; it never advises.
How we work together
Most London clients start with a fixed scope build covering one website and one messaging channel, then move to a monthly retainer for content updates, new channels and transcript reviews. Larger organisations with several brands or departments sometimes take a dedicated team for a longer programme. All of it runs remotely: discovery, demos, testing sessions and launch reviews happen by video call with whoever owns customer service on your side, and a shared test link lets your staff try the chatbot from their own phones between sessions.
A defined build with a deadline
We agree what the product has to do, write it down with a price and a date, and build exactly that. Changes are priced before they are made, not discovered on the invoice.
A product that keeps evolving
Named engineers who work your hours, join your standups and stay on the codebase. You direct the backlog; we own the quality of what ships.
Something live that needs looking after
A monthly block of hours for improvements, fixes and the small requests that otherwise wait months. Unused hours carry for a quarter.
Technology
We choose the language model, hosting region and integrations to suit your data rules and the tools your team already uses. That can mean models from providers such as OpenAI, Anthropic or Google, retrieval over your own content, and connectors for help desks like Zendesk or Freshdesk and CRMs such as HubSpot.
01Mobile
02Data and cloud
03Marketing and tracking
It depends on your content, channels and integrations, and we confirm it in writing after discovery.
A website chatbot answering from tidy, current content is the quickest case. Timelines grow when the knowledge base needs rewriting, when several messaging channels and languages are involved, or when the bot must read live data from a booking system or CRM. We estimate after reviewing your enquiries and content, then write the milestones into the scope. Testing with your own staff is part of that plan from the start, because it is where most of the useful fixes come from.
Schedule a free consultation and get a written recommendation, with costs, before you commit to anything.
Get in touchPROVEN RESULTS
Swift
Kotlin
Flutter
React Native
PostgreSQL
MySQL
MongoDB
AWS
Azure
Docker
GA4
Google Tag Manager
Search Console
Meta Pixel
Looker Studio