For the last one year and a half, I've been working as a UX Designer with a focus on conversational interfaces - aka chatbots.

Since then, I had the chance to create and develop "virtual assistants" for over ten clients and several projects, from discovering use cases and personas to selecting and tailoring the best componentes and copies to the user flows.

In this period, I worked in two main fronts:
Consultant - in close contact to clients, understanding their businesses challenges and creating experiences to satisfy their users' needs.
Product designer - as part of a bigger Research and Development team, I took the B2C step resonsibility in one of our company's most disruptive products.

Acting as a consultant meant some common challenges, such as:
- existing product (legacy) with flaws and biases to be corrected as well as new features to implement;
- tight budget - meaning more creativity needed to guarantee good error handling avoiding unnecessary human service or acquirign users without paid notifications;
- large number of stakeholders, from operational to strategic levels;
- top-down decisions with little room for adaptations;
Some of my main activities were:
- understanding the product business rules and uncovering its details and intrincacies (use cases, data handling from APIS, etc);
- "translating" the documentation to my company's convention;
- applying the design system to the conversational flow;
- keeping track of the product performance and suggesting improvements based on the data;

Below: conversational user flow in a client's documentation (before).

Below: a client's conversational user flow after I documented it in macro steps and following information architecture conventions, with contrasting colors to the flowchart's elements and a caption on the right side.

Below: left (before) - product's business rules documented by the client in a simple spreadsheet, without visual cues or highlights; 

right (after): documentation of business rules with visual symbols, using colors and typography to better inform the squad (and even the client!).

Below: left (before) - chatbot copy documented in a spreadsheet, without indication of component or interaction possibilities;

right (after): copy documented in its own component, using adequate formatting.

Below: left - chatbot high fidelity prototype on Figma with scrollable conversation and interactive menu; right - chatbot interactions documented on Figma;

Below: product UX documentation covering development start date and main changes, tone and voice, users profile, context and main goals, features, main improvements and pain points, amid others. 

Below: user journey and heuristics analysis.

Below: users track analysis and documentation.

Below: A/B test with hypothesis, key indicator and success metrics, codesign versions  documentation.

Below: left to right - questionnaire, product requirements mapping, user semantic research