← Back to Blog
Using GPT-4 to Automate Lead Qualification in Django
# Using GPT-4 to Automate Lead Qualification in Django
My client's agency was drowning in leads — 200/day, most of which were garbage. The sales team was spending 3 hours every morning manually sorting them. I built an AI pipeline to fix that.
## The Architecture
```
Lead submitted → Django view → Celery task → GPT-4 scoring → CRM update → Alert if high-value
```
## The Scoring Task
```python
@shared_task
def qualify_lead(lead_id):
lead = Lead.objects.get(pk=lead_id)
prompt = f"""
Score this lead 0-100 and categorise as HOT/WARM/COLD.
Name: {lead.name}
Message: {lead.message}
Package interest: {lead.package}
Respond with JSON: {{"score": 85, "category": "HOT", "reason": "..."}}
"""
response = openai.chat.completions.create(
model="gpt-4",
messages=[{"role": "user", "content": prompt}],
response_format={"type": "json_object"}
)
data = json.loads(response.choices[0].message.content)
lead.notes = f"AI Score: {data['score']} ({data['category']}) — {data['reason']}"
lead.save()
```
## Results After 30 Days
- Manual review time: 3hrs → 20min/day
- Conversion rate: 8% → 23%
- Revenue attributed: +$12,000/month