Bid Brokers

Investing in auctioned property demands speed on three fronts: knowing what the property is worth, closing the contract without friction, and operating at low cost. We built the engineering for all three.

Model-based pricing

• A machine learning model trained to price registered properties, served through an API: the team registers, the model returns the value;
• Market data collected from real estate portals and enriched via BigQuery.

Closing automation

• Contracts filled automatically from deal data, with amounts and dates spelled out by a dedicated service;
• Digital signature integrated via D4Sign, with confirmation flowing back into the pipeline.

Operations

• Directus back office with permissions applied by script;
• Serverless Python microservices (AWS Lambda) and infrastructure described in Terraform: cost proportional to usage.

The engineering result: a pipeline where a property comes in raw and leaves priced, contracted and signed, with a lean team running the process.