Desay SV, one of the leading mobility technology companies, is committed to being the preferred partner in mobility transformation. The AI Application (AA) Team explores frontier AI technologies, with this role dedicated to advancing applications in large language models (LLMs).

Project Background:

With the rapid advancement of large language models, optimizing and fine-tuning them for specific applications has become increasingly essential. This project aims to leverage state-of-the-art LLM techniques to enable automated documentation generation tailored to specific requirements, significantly boosting productivity and efficiency.

The primary objective of this internship is to familiarize the student with the concepts, methodologies, and tools of LLM application. The intern will assist in:

• Creating pipelines for requirement-specific document generation.
• Fine-tuning local LLMs for increased accuracy and efficiency.
• Utilizing chain-of-thought and prompt engineering to improve model outputs.
• Collaborating with teams to facilitate smooth deployment in production.

The scope of the internship will typically include the following tasks:

• Fine-tuning Local LLMs: Optimize and fine-tune the local DeepSeek R1 large language model to achieve enhanced performance.
• Data Engineering: Construct and curate high-quality training datasets tailored to language model applications.
• Chain-of-Thought Design: Develop and implement innovative chain-of-thought frameworks to improve model reasoning capabilities.
• Prompt Engineering: Design and refine prompts to effectively steer model outputs and improve response quality.
• Research & Experimentation: Collaborate with cross-functional teams to explore novel ideas and cutting-edge methodologies in AI and natural language processing.

Requirements:

The student should possess the following skills:

• Programming: Proficient in Python and/or C++.
• Environment Management: Experience with Conda and/or Docker.
• LLM Knowledge: Familiarity with the characteristics, principles, and architectures of large language models.
• Practical Experience: Previous experience with local deployment and fine-tuning of large language models is a plus.
• Project Background: Relevant projects or internship experience in AI, machine learning, or NLP are advantageous.