The challenge
ACE Australia's consultants were busy and its client list was growing. A steady stream of early-stage enquiries came in, and each one needed a quick, helpful reply. ACE wanted to answer them without losing the personal guidance it is known for.
What we built
We built Acey, an assistant that runs on OpenAI's GPT-4 model. Acey is ACE Australia's version of Prospect Copilot, our AI assistant product. We set it up for Australian education and immigration, and taught it ACE's processes and culture.
Acey talks to early-stage leads and gives general advice. When a lead is ready, it points them to one of ACE's consultants. During the conversation it captures details such as names and emails and writes them straight into the CRM. It also scores each lead and routes it to the most suitable person on the team.
Key features
- Built on OpenAI's GPT-4 model
- Tuned to Australian education and immigration and to ACE's own processes
- Gives general advice, then hands over to a human consultant
- Captures names and emails straight into the CRM
- Scores each lead and routes it to the most suitable operator
Results
Early-stage leads now get a prompt first reply that fits their question. About 80% of first-contact enquiries are handled without staff, and the median first response is about 60% faster. Both are estimates, not audited measurements. Because Acey is connected to the CRM, capturing details and assigning leads take no manual steps, and consultants have more time for the more complex client questions.
Figures marked ~ are estimates, not audited measurements.
Why it matters
Acey has a narrow job: first contact, general advice, capturing details and routing. When a conversation needs a person, it hands over, so consultants keep the work that needs their judgement.