Rotoplas
Role
Digital Product Designer: (Concept design, UX design, UI design, Wireframing and prototyping, Front end UI, User test)
Tools
Figma, Rails, Tailwind, AI
About the project
The client
Rotoplas es una empresa lider en soluciones de almacenamiento y filtración de agua en México con sucursales y fabricas de quipos especializados dentro y fuera del páis. recientemente interesada por la implementación de la AI dentro de sus procesos y áreas organizacionales.
The Problem
Due to being a large company with multiple branches and factories distributed both domestically and internationally, it faces daily challenges in reconciling data and information from each of these locations. This affects various areas of the organization, specifically inventory, sustainability, and the legal department, slowing down decision-making and the continuous improvement of its internal processes.
The Background
El equipo de Icalia Labs se reunió durante una semana con los diferentes equipos de Rotoplas para registrar e identificar las puntos de dolor en los diversas áreas y procesos. Lo anterior partiendo desde el interes del cliente en la creciente ola de hype al rededor de la IA y cómo esta se podría implementar para resolver sus problemas organzacionales. Durante este ejercicio no solo se identificaron dicho puntos de dolor sino también se recopilaron ideas para la solución ideal de los mismo, obtenidos desde su perspectiva cómo usuarios e Icalia Labs cómo expertos en la materia de IA.
Posteriormente me uní al equipo encargado del proyecto para ayudarlos a transformar la información recaudada en 2 pruebas de concepto a lo largo de 4 semanas que puedan solucionar los problemas de las diversas área con la ayuda de la inteligencia Artificial.
Ideating the solutions
Understanding the client expectations
From the beginning of the project, my goal was to create a simple experience based in the already know AI experiences (chat based platforms), with the objective of efficient development time and a seamless interaction from the users trough the 2 solutions, but once i started analyze the problems in deep, i identify that the ideal output for the 2 solutions where way different, discarding this approach and starting building from the outputs to the user interactions.
At the beginning of every solution we expend 3 days of the 2 weeks to talk again with the product owner from the area of the solution, this was so I could fully understand the perspective of future users and their problems, and in those first two days we were able to jointly devise a digital solution for their pain points.
Solution 1: inventory management
Ideation
For the first solution, we worked extensively with the 'Inventory' area, where the main pain point was the data reconciliation process between the management tools they use (SAP, Salesforce, and an internal Excel sheet in warehouses). The process to validate any in and out of equipment for their in-house water filter solutions was done manually by almost 10 people across the country’s warehouses. This led to a slow validation process and working with information that was at least a day outdated.
By focusing on the expected outcome, we used the final document of the manual reconciliation as our basis, ensuring we did not disrupt other processes dependent on this activity. Subsequently, we added relevant information that AI could detect from the automatic reconciliation and defined how this information could be collected easily and quickly.
Design | 1, 2, 3, Reconciliate
I ended with a 2 screen experience due the only action possible was just ‘upload data’ and ‘view data’, overcomplicate the experiences would be unnecessary.
The ‘upload view’ was thought to be a straightforward form, while specifying well the needed documents and parameters to be analyzed in a steps flow presentation. At the beginning the idea was to just can conciliate a warehouse per request, but after the easy development and user feedback, we open the possibility to reconciliate multiple warehouses at the same time, so this needed an improvement in the experience
Design | Accessible information
The ‘view data’ screen was more straight forward due their only use was tho display information, and the options to switch between the uploaded warehouses. This screen was a conjunction of tables with the conciliated data, indicators of variations in the info and which sources present those variations. All been available for download in XMLS and CVS so they can work with the right data.
Results
The inventory reconciliation team completed a POC that truly addressed their main reconciliation issues, reducing the wait time from 2 days to just 1 to obtain real data (due to their continued use of SAP and the consumption data being available only a day later). Additionally, there was a reduction in the human resources required for the reconciliation of all the warehouses in operation.
They also expressed satisfaction with the overall experience of the POC and the direct approach, feeling a general sense of "ease" and practicality for a task that previously felt overwhelming and repetitive, allowing the team to better focus their efforts on other tasks in the area.
-50%
of staff and time
to reconcilaite data
Solution 2: legal
Ideation
The Legal area solution was the more easy on paper to implement, they need a way to keep scope of the statues of the different concession and license for every factory alongside the country. And a way to quick translate those status into a visual graph. This solution actually didn’t require a AI integration in a standard development, but we use it to efficient time in the development and the status tracking system to be more acquired and to provide insights and recommendations about the current status of the factory licenses.
Factories licenses tracker
I focused on the important of the quick and easy status view, treating every license as a singular product inside every factory, designing an status alert system to be the main feature in the solution.
Starting with a visualization of the individual factories, and their overall number of alerts, also implementing the feature of add new factories, this way they can keep control of all the factories they need in the future.
Keeping an eye in statuses
Inside every factory the user is received by the main notification center section where users can visualize more in deep the needed actions and recommendations of the selected factory
Ready to share chart
Following the user finds the validation over time chart, with a designed interaction for month per month details, ready to be downloaded by the use in PNG format
Licenses details and management
Finally every factory has a collection of licenses where the user can quick visualize the number of licenses, their status and other stakeholders requested documents if they are attached. in this section the user can also update the licenses document
Within each license, the user can modify the documents associated with the licenses (license, license application, and payment receipt for the license). The main function supported by AI reads the document and automatically provides the start and expiration dates; however, we allow the user the freedom to modify this data if the AI made any reading errors. Lastly, the stakeholders requested the ability to save up to the last 3 licenses uploaded for administrative purposes, so we provided a section for this where the user can download them.
Challenges and overcomes
Entering in the middle of the process
I realized that missing the week-long workshop with the client made me feel disoriented about the real problem in each area of the client. Despite having the information in writing and “filtered” when it comes to user experience, I prefer to have first-hand information from the client, paying attention to the expressions and language with which they communicate the problem or situation they are going through, it helps me prioritize the functionalities, interactions and necessary steps in a better way than just reading them. Because of that, the meetings at the beginning of each POC were of great need and help for the final ideation of the solutions. I am left with the thorn that that time could have been used to refine the final product. However, I am satisfied with the platforms we built with the available time.
Complexity in the process
Being the projects for a company dedicated to the manufacture, this was my first approach to a group of big amount of data for areas that I didn't know, and process that i haven’t event listen about, and being my role the designer of how the users and the AI will interact with the information made me feel overwhelming at the beginning but once i started learning about the process and how the user interact with the amount of data and the important parts to them, also with the help of the AI to read, translate and explain me the content of the files, i started to get clarity abut wht really matter for the final solution and how should be.
Working with POC
For me, this was the first time i have worked on the development of probe of concepts, this implicate to change my methodology and process troug the design process, i usually try to expend time creating different designs for the same feature and ask for feedback and run small test, before sharing any final art to the development team, this usually takes the 1 to 2 initial weeks while the developers setup the project and get all the needed data, but this time we only had 2 weeks for the entire design and development, and i spend 3 of those days learning more about the project and user needs and specifications. so for this time i have to be confident in my design decisions and experience to go with the more optimal flows and designs.
To overcome this challenge i shorten my process to only 3 days, working with only 2 design options and selecting one by me, using research and compare from similar experiences in the market and also scarifying aesthetic aspects. One thing that help me in my decision making was talk to my team and the business unit, understanding better how a POC works and its finality calm my mind and my process, prioritizing functionality over aesthetic.
