Chatbots their customers build themselves

A webchat platform for a review management company, letting its customers build and plug in their own chatbots without an engineer.

Talk to an engineer Data engineering & ML

Review management Client

Data engineering & ML Practice

July 2024 Published

01 — The challenge

Where it started.

The platform wanted to give its existing customers a chatbot, which meant the customers had to be able to build one — not file a request for one.

Anything requiring configuration by a developer would have reached a fraction of them.

02 — What we built

The work itself.

Python and Flask behind it. Training and chatbot setup handled as a service rather than per customer.

OpenAI, grounded in their own pages. Replies generated against publicly available content, with the client’s GraphQL integration supplying the data.

Webchat, with the look configurable. Colours, welcome and fallback messages set per chatbot, with the backend API behind it.

A bot that plugs straight in. Automated creation of a chatbot the customer drops into their own portal, capturing enquiries through GraphQL where the conversation calls for it.

03 — Results

What the client has now.

Customers create several chatbots and place them on their own sites. Setup measured in minutes, with creation, training and responses automated behind it. Deployment that needs no engineering knowledge from the person doing it.

The rest of this study

What we built, what it produced, and the stack behind it.

We share the whole write-up — we would just like to know who is reading. One form opens every case study on the site for the next 30 days.

If this is your problem too

Start with a fixed-price pilot.

Two weeks on a real piece of your backlog, priced before we begin. You keep the output whether or not you continue.

Related work

Consumer credit · Document automation

The manual redaction step, removed

Healthcare · Synthetic data