Product Review: CodiumAI
Generate tests and elevate code integrity
--
Disclosure: I am a strategic advisor to CodiumAI, and a former testing lead/systems integration consultant who shares their mission and vision, and support the solution for pain points and unmet needs their product addresses.
Software bugs are the worst. Besides leading to incorrect errors, unexpected results, long nights and frustrated developers, bugs pose a significant financial strain: bugs cost United States firms $607 billion in 2020, according to the Consortium for Information and Software Quality (CISQ), up 21.4% from 2018. Incorporating second- and higher-order downstream effects, overall associated losses ballooned to $2.08 trillion, accounting for operational failures, technical debt, legacy systems, unsuccessful projects, business continuity impacts and disaster recovery. What then, can be done about software bugs? Granted, the answer lies in testing, though this poses its own costs across time, energy, and money: in the rush to ship against deadlines and comply within budget and scope, critical edge cases can be missed, amidst increasing frustration and tedium.
Yet, on the bright side, we have arrived at a massive inflection point in technology in general, and software development in particular, with the release and mass adoption of large language models (LLMs) ChatGPT and GPT-4. With the increasing accuracy, efficiency, and relevance of succeeding versions, developers have leveraged these powerful, novel technologies to increase productivity and profitability, reduce costs, and eliminate manual and/or rote tasks. The natural question, given these developments: what if the scale and scope of artificial intelligence could be exploited and extrapolated onto software testing?
Enter CodiumAI.
Powered by TestGPT, CodiumAI generates tests to catch bugs and check code across Python, JavaScript, and TypeScript, within the integrated development environment (IDE), through an interactive user experience, for greatly reduced testing (and overall development time), while still maintaining excellence across unit tests and use case ideation.
Usage is delightful, elegant, and straightforward. Available as an extension for both VSCode and JetBrains IDEs, CodiumAI can be effortlessly installed, and easily deployed right away. On the CodiumAI homepage, navigate to “Get CodiumAI”, visit the appropriate link (VScode or JetBrains), download and install the desired version: sign in, authorize, and get started! One-click “generate tests” above the relevant class or function will generate a variable set of suggested tests for miscellaneous use cases. Users have access to an exhaustive array of features and functions: review results, exclude/include tests, modify their test style, add test examples, starting tests, input data, modify frameworks, regenerate tests, edit test objectives, request additional tests, copy and save tests.
Since CodiumAI parallelizes and chains multiple calls, an eclectic, discrete corpus of tests can be generated, with broad code context for prompts, and the ability to fine-tune the suite with references to existing tests, input samples, and much more. CodiumAI is devoted to security, privacy, and compliance, and analyzes the minimum of necessary code, with full SSL data encryption, and System and Organization Controls (SOC) 2 compliance certification in process. These functions will altogether simplify and elevate code integrity, while eliminating the need for boilerplate junk code tests. Today, CodiumAI launches its open beta and invites the developer community to take part in the next revolution in code logic checking and code integrity improvement. Build faster with fewer bugs.
Visit CodiumAI today!
P.S. CodiumAI is LIVE on Product Hunt! Every vote and comment counts. Your support is greatly appreciated. Much obliged and many thanks! 🙂
