I led the design of Gluecharm, an AI-driven Technical Analyst tool that converts product or feature concepts into detailed technical specifications.
This initiative aimed to enhance collaboration between product owners and development teams, streamlining the review process and accelerating development cycles.
Context and Main Challenge
Gluecharm was a side project I developed between 2022 and 2023, inspired by the rise of generative AI and the complexities faced by technical teams when creating high-level documentation within software development.
The main goal of the project was to create a tool that simplified these processes, automated repetitive tasks, and provided an efficient system for structuring and generating technical documentation, regardless of the user's experience level.
You can find more information about Gluecharm at www.gluecharm.com.
Research Process
The first step was to collaborate with the product team to develop a functional MVP. During this stage, we defined execution timelines and established clear priorities to ensure the project's success.
User Interviews
To better understand the needs and challenges of potential users, I conducted several focus group sessions with both technical and non-technical profiles. These sessions allowed us to identify the main issues associated with generating complex documentation for software development, such as lack of automation, excessive time spent, and insufficient clarity in document structure.
Focus Groups: Key Insights
Based on the gathered insights, we identified three main areas for the tool to focus on:
Structured report generation.
Task automation.
Improved data analysis processes.
Shadowing with Technical Roles
I complemented the focus groups with shadowing techniques, observing how key roles like product owners and analysts performed their daily documentation tasks. This revealed several critical pain points, including:
Excessive time spent on documentation.
Lack of a standardized structure for documents.
Complexity in creating tasks in tools like Jira.
This research enabled us to design a tool centered on real user needs and adapted to their workflows.
Problem Definition and Proposed Solutions
Main Problem
After analyzing the research results, it became clear that technical documentation faced two major challenges:
Complexity for non-technical users, which hindered their participation in the process.
The time and effort technical roles spent structuring and updating documentation.
Proposed Solutions
We designed a tool based on simplicity and usability, allowing any user to interact with the AI to generate complex technical documentation. The solution was structured around a narrative in which the main user answers three key questions, and the tool handles the rest:
What problem are you trying to solve?
How do you want to solve it, or why?
Who is your user?
With this information, the tool generates:
Suggested names for the product.
Identification of problems and proposed solutions.
User definitions with customizable preferences.
Use cases, scenarios, technical diagrams, and data models.
Additionally, it allows users to convert these cases into user stories, with AI assistance to enhance and clarify the documentation through guided, simple steps.
Technical Development and Design
To ensure the tool was functional and scalable:
I created wireframes detailing each step and user flow, ensuring all aspects aligned with identified needs.
I implemented a design system based on the Tailwind library, leveraging pre-built components to streamline design and development.
Results and Learnings
Results
I designed a stable, user-friendly platform tailored for both technical and non-technical users.
The tool successfully automated key tasks, significantly reducing the time and effort spent on technical documentation.
We achieved a consistent and scalable user experience by leveraging the Tailwind design system.
Learnings
Interdisciplinary collaboration: Working with technical and non-technical roles provided a comprehensive perspective, crucial for designing a versatile solution.
Optimization through design: Integrating a design system from the start enabled the development of a functional and visually consistent tool.
AI applied to real-world cases: This project reinforced the importance of simplifying complex processes through AI-driven tools.
