Testing | AI | Automation

Building a Testing Strategy

case-img

Context

A startup leverages AI systems to integrate them within SDLC (Software Development Life Cycle) process. The application communicates with LLMs like GPT, Claude, Gemini and automates regular user tasks like ticket creation and code writing.

Challenge

Since the product hadn’t got any testing processes apart from manual verification from the development team, bug rates were increasing daily and produced negative feedback. We needed to establish such testing processes, which should be shaped to the current product budget assuming it is a startup and there should not be a lot of bureaucracy here

Solution

Present the testing strategy to a customer and implement it. Prepared testing pyramid with several levels to automate. Prepared documents with manual test cases from scratch and set priorities. Communicated any sort of bugs or unexpected behaviour immediately to a team. Automate manual critical test cases.

Results

The release cycle became faster and smoother. QA team handles all testing activities and set them as a quality gate. The bug rate decreased up to 50%.