It is a well-known fact that developers should not be the ones to test the code they build. They usually dislike the task and, as a best practice in auditing, the work should be verified by someone else.
Code testing, whether it’s usability, language-specific, or end-to-end testing, has attracted the attention of several generative AI startups. Some notable ones include Momentic, Antithesis, CodiumAI, and QA Wolf, all raising significant funding. Nova AI, a year-old startup, is another player in this space, focusing on mid-size to large enterprises with complex code-bases.
Nova AI sets itself apart by automating test generation using GenAI. Their tool is designed for continuous integration and delivery environments where ‘heavy user experiences’ require minimal downtime. The founding team, with experience at tech giants like Google and Meta, understands the importance of efficient code testing.
While many AI startups rely on OpenAI’s models, Nova AI has chosen to use alternative open-source models like Llama and StarCoder. This decision, while driven by privacy concerns of enterprise clients, has also proven to be cost-effective and efficient for their specific testing needs.
By leveraging open-source models and deploying their own customized tools, Nova AI aims to provide targeted solutions for code testing without compromising on data security. This approach has garnered interest from large enterprises looking for reliable and privacy-focused AI solutions.
