LLM Engineer for Voice Sales Objection Handling

We’re building a voice-based sales AI to handle real-time objections during outbound phone calls. An experienced LLM engineer is needed to set up and host Mistral 7B, integrate a RAG system using objection-handling data, and fine-tune the model with transcribed call data.

Responsibilities:
- Set up a vector store (ChromaDB, Pinecone, or Weaviate)
- Implement a retrieval + prompt injection pipeline using LangChain or LlamaIndex
- Fine-tune Mistral 7B with proper JSONL formats.
- Host the model and ensure it meets latency needs.

Requirements:
- Strong experience with LLMs and fine-tuning.
- Familiar with RAG frameworks and hosting on RunPod or similar platforms.