model.py

from transformers import pipeline
from schema import PreCallSummary

class PreCallSummaryModel:
    def __init__(self):
        # Load FLAN-T5 locally instead of via HF Inference API
        self.generator = pipeline("text2text-generation", model="google/flan-t5-base")

    def generate(self, context: dict) -> PreCallSummary:
        prompt = self._build_prompt(context)
        result = self.generator(prompt, max_length=512, do_sample=False)
        summary_text = result[0]["generated_text"]

        return PreCallSummary(summary=summary_text)

    def _build_prompt(self, ctx: dict) -> str:
        return f"""
        You are a helpful assistant. Generate a professional pre-call summary
        based on this Salesforce account context:

        {ctx}

        Format the output as a concise summary suitable for a sales rep.
        """