I think the term
because our customer isn't a software engineer and we know that our data model will change many times
Is a reality but is not the whole reason of data model change in many times!
BTW some suggestions may be helpful
Consider business analysis a key point and do continues business analysis
More and more investing on analysis will cause less cost on development and change management
Have a specific talented analyst as a representative positioned in the customer office.
Have a specific representative of customer accountable for analysis and development team.
Document the customers business process, talk about them with the customer. Document the negotiation results.
Design and change management
Never design the database alone. Brainstorming team member, invite customer representative, invite business analyst, have a DBA in design team, have a Change-Manager role.
Technology
Have a look on technology trend (may using a document-based database will be helpful)
Have a flexible framework and architecture in service of business software production and not in reverse!
Changes will come ... be ready
Changes will come, all above efforts are targeted to optimized the solution to cost less when change happen, they will not prevent the changes.
You need to have an acceptable mechanism to bill them to the customer
ITIL's Service-level management, SLA and OLA will be a guide line.
So the question answers:
1) All above will help on the problem. Continues interactive analysis first, next design and develop in iterations, using template approach.Changs will come and they will cost.
2) Depends on technology and frameworks, there will be tools (I don't know EF) just don't insist on a specific platform or library.
Hope this helps.