Just Now Possible Building Tendos AI: How an Agent Swarm Turns Construction Emails into Quotes
49 snips
Jan 15, 2026 Daniel Kappler and Matthias Hilscher from Tendos AI dive into the transformative potential of AI in construction. They discuss how focusing on specific products like radiators helped them prove their model's value before expanding. The conversation highlights the tendering chain's inefficiencies and the role of document extraction from PDFs, including managing vast amounts of data. They emphasize the importance of both automated and human reviews for quality control, and elaborate on how customer feedback has driven product development.
AI Snips
Chapters
Transcript
Episode notes
Tendering Chain Is Deeply Inefficient
- The tendering chain in construction is a multi-party, high-friction process that creates daily inefficiencies.
- Daniel Kappler says projects can take months due to many parties and manual handoffs across manufacturers and wholesalers.
Home Renovation Bid Experience
- Teresa Torres describes her home plumbing and GC experiences to illustrate how fragmented bids and subcontracting feel to end customers.
- She contrasts a detailed bid (clear scope) with a five-line bid that left her worried about hidden costs.
Requests Are Semantic, Not Structural
- Incoming requests arrive as unstructured text and attachments, requiring semantic extraction to prioritize and scope work.
- Daniel Kappler explains Tendos extracts positions, functional descriptions, and context to prepare a quote for human review.