AI/NLP Engineer Needed - SMS Automation System (RoBERTa + AWS Lambda)

What We Need

We have begun building an AI for our car buying company that responds to customers in regards to their vehicle. We are looking to complete our SMS automation system for a car buying business. We have RoBERTa/Lambda infrastructure started and we need you to finish it.

• *Tech Stack:**
• RoBERTa for intent classification
• AWS Lambda
• Twilio SMS
• Redis
• Zoho CRM

## The Job

Build an AI system that

1. Reads and responds to customer SMS messages

2. Classifies their intent (accept offer, negotiate, not interested, etc.)

3. Extracts key info (Addresses, VINs, Title Pictures)

4. Generates smart responses

5. Updates our CRM automatically

We have a buying team who gives a customer a quote initially, then we want AI to followup, negotiate and close deals.

We have some of the basic handled, but it's still kinda clunky. For example, it gets the intent of the customer confused. It could followup with a customer asking if they want to take offer, but the customer will respond "still looking" and the AI will respond with asking them who the company is they sold it to which doesn't make any sense. The AI is having trouble figuring out the intent of the customer and responding incorrectly in some cases.

## Requirements

• *Must Have:**
• Production experience with RoBERTa, BERT, or similar transformers
• AWS Lambda deployment experience
• Built intent classification systems before
• Python expert
• *Nice to Have:**
• Twilio API experience
• Zoho CRM integrations
• NER (Named Entity Recognition)
• Redis

## Deliverables

• Fine-tuned RoBERTa model (90%+ accuracy)
• AWS Lambda functions deployed
• Full SMS → AI → CRM pipeline working
• Documentation
• *Start proposal with:** "I've built SMS intent classification systems before"

No generic proposals. Show us you've actually done this. I have attached the architecture we're looking for.