BPS (Breaking Point Systems) is a L23/L47 network traffic generator. It can replicate realistic traffic patterns at very high loads and can also simulate security scenarios like malware, DDoS attacks and so on.
We run periodic regression test suites to catch bugs early. Regressions take a long time to run (days) and, because of the large number of tests, any regression run requires a significant effort from our QA engineers, which have to painstakingly go through all the failed test cases in an attempt to narrow-down the root cause of the failures. Automating this process using AI would free up significant QA resources for other tasks.
You’ll work on helping to automate part of the regression analysis process using AI (LLMs):
🔹 Uses AI/LLMs to analyze test results from the regression suite. 🔹 Analyze data from multiple sources (regression framework logs, application logs, system logs, recent changes from GIT commits etc.) 🔹 Produce a report summarizing the main issues observed during the regression and their potential causes. 🔹 Identify and flag changes from GIT that may be responsible for the observed failures.
Since the volume of data is non-trivial, feeding it all to an LLM seems unrealistic. We expect some form of agentic system to be required to intelligently navigate through the data.
What you will gain:
- Apply AI (LLMs) to solve real-world problems
- Experience developing an AI system that requires efficiently parsing and analyzing vast amounts of information
- Experience working with agentic systems
Skills required: LLMs, python, agents
