HomeLaunchesAvoice
70

Avoice: The AI workspace for architects

Harvey for Architecture

Avoice is the AI workspace for architecture firms, using AI agents to automate the operational work that takes time away from design. The platform understands messy project data - specifications, schedules, codes, drawings, materials, and notes - and turns them into structured, usable outputs across documentation, coordination, QA/QC, and research.

Instead of architects manually stitching together PDFs, spreadsheets, and folders, Avoice’s AI agents act as digital teammates that can write specifications, check compliance, coordinate schedules, and surface project knowledge across a firm’s past work.

Firms using Avoice already manage $300M+ in active projects on the platform.

Problem

Architects spend an enormous amount of time on work that is critical but painful: writing specifications, coordinating schedules, checking codes, redlining documents, and keeping project knowledge consistent across teams.

This work is:

  • Fragmented across PDFs, spreadsheets, emails, and folders
  • Largely manual, repetitive, and error-prone
  • Poorly supported by modern tools (most software optimizes drawings, not documentation)

The architecture industry is also a massive ($300B+) but historically underserved market: highly fragmented, dominated by small and mid-sized firms, and reliant on legacy software built decades ago.

Most incumbents focus on drawings or rigid document systems that require heavy customization and consulting to work for each firm. As a result, many architecture practices still rely on patchworks of spreadsheets, PDFs, and disconnected tools to manage documentation and project knowledge.

As a result, firms lose time, consistency, and institutional knowledge across projects , especially as teams scale or turn over.

Approach

We built Avoice around how documentation is actually produced inside architecture firms — specs, schedules, coordination, and QA — not just drawings.

Avoice acts as a shared workspace where project information lives in one place and stays connected as it evolves. Instead of rewriting the same information across different documents, teams can reuse, reference, and update it consistently.

Historically, architecture software struggled because every firm works slightly differently. Customizing workflows and documentation systems for each practice required expensive consulting and was difficult to scale. AI changes this equation.

With Avoice, AI agents can adapt to a firm’s specific standards, templates, and workflows, helping generate specifications, coordinate schedules, check for coordination issues, and support QA while still keeping architects in control of decisions and design intent.

Our goal isn’t to change how firms practice architecture, but to reduce the manual overhead that slows teams down and introduces avoidable

https://youtu.be/k0C4puetlV8