Finding the Perfect Mid-Size Model That Destroyed Mid-Range Competition

import ollama
import pandas as pd
from datetime import datetime

class SMEAIAssistant:
    def __init__(self):
        self.model = 'mistral-nemo:12b'

    def analyze_business_document(self, doc_path):
        """Analyze business documents with GDPR compliance"""
        with open(doc_path, 'r', encoding='utf-8') as f:
            content = f.read()

        prompt = f"""
        As a European business analyst, analyze this document:

        {content}

        Provide:
        1. Key business insights
        2. Action items
        3. Risk assessment
        4. Compliance notes (GDPR)

        Format as structured business report.
        """

        response = ollama.chat(
            model=self.model,
            messages=[{'role': 'user', 'content': prompt}],
            options={'temperature': 0.3}  # More consistent for business use
        )

        return response['message']['content']

    def multilingual_customer_response(self, customer_msg, language='auto'):
        """Handle customer inquiries in multiple EU languages"""
        prompt = f"""
        Customer message: {customer_msg}

        Respond professionally in the same language as the customer.
        Follow European customer service standards.
        Be helpful, concise, and culturally appropriate.
        """

        return ollama.chat(
            model=self.model,
            messages=[{'role': 'user', 'content': prompt}]
        )['message']['content']

# Usage example
assistant = SMEAIAssistant()
report = assistant.analyze_business_document('quarterly_report.txt')

# Deploy as REST API service
from flask import Flask, request, jsonify
import ollama

app = Flask(__name__)

@app.route('/api/document-analysis', methods=['POST'])
def analyze_document():
    """GDPR-compliant document analysis endpoint"""
    data = request.get_json()

    # Log for audit trail (GDPR requirement)
    audit_log(request.remote_addr, 'document_analysis')

    response = ollama.chat(
        model='mistral-nemo:12b',
        messages=[{
            'role': 'user',
            'content': f"Analyze this European business document:\n{data['content']}"
        }],
        options={'num_predict': 1000}
    )

    return jsonify({
        'analysis': response['message']['content'],
        'timestamp': datetime.now().isoformat(),
        'gdpr_compliant': True
    })

@app.route('/api/customer-support', methods=['POST'])
def customer_support():
    """Multi-language customer support"""
    data = request.get_json()

    response = ollama.chat(
        model='mistral-nemo:12b',
        messages=[{
            'role': 'system',
            'content': 'You are a helpful European customer support agent.'
        }, {
            'role': 'user',
            'content': data['message']
        }]
    )

    return jsonify({
        'response': response['message']['content'],
        'detected_language': detect_language(data['message'])
    })

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)