At Cadence, we hire and develop leaders and innovators who want to make an impact on the world of technology.

This project entails the evaluation of LLM and a RAG pipeline in customer support scenarios.

This project involves:

• Providing user prompts to the Cadence JedAI RAG platform to generate answers to customer questions and evaluation of the quality of  results.
  • Organize the responses and high and low quality results
  • Try to find categories of questions/problems that have higher probability of success then others
• For low quality responses attempt to debug the pipeline and deep dive into problems like
  • document ingestion process (document chunking, hierarchical data organization, etc.)
  • problems on insufficient contexts for the LLM to generate responses
  • problems in scripting/implementation of RAG pipeline itself
• At the end of the project create a report on feasibility of RAG/LLM pipeline in Tensilica Custer Support scenarios.

Requirements:

The ideal candidate will have experience with projects involving Text/Speech neural networks and LLMs.

Currently enrolled in MS or BS majoring in CS, CE, EECS, or equivalent.

We’re doing work that matters. Help us solve what others can’t.