Bella led the HubSpot automation product team, where she launched many workflow automation products in Marketing, Sales and Service Hubs. She was the Head of Product and Growth at Dance, where she scaled the member base to 10k+ and built the tools for ops. She graduated from UChicago with a BA in Economics.
Bella Wu
Founder
Bella led the HubSpot automation product team, where she launched many workflow automation products in Marketing, Sales and Service Hubs. She was the Head of Product and Growth at Dance, where she scaled the member base to 10k+ and built the tools for ops. She graduated from UChicago with a BA in Economics.
Vincent Garrigues
Founder
Vincent has 12+ years of experience leading engineering teams.
He joined SoundCloud as the 2nd iOS engineer and worked his way to director of engineering, overseeing the media streaming, player, and core listener teams (iOS, Android, Web, Backend, C++).
As principal engineer at Dance, he worked on everything, from customer facing apps to internal web and mobile apps for the operations team.
Vincent Garrigues
Founder
Vincent has 12+ years of experience leading engineering teams.
He joined SoundCloud as the 2nd iOS engineer and worked his way to director of engineering, overseeing the media streaming, player, and core listener teams (iOS, Android, Web, Backend, C++).
As principal engineer at Dance, he worked on everything, from customer facing apps to internal web and mobile apps for the operations team.
Caterina Kiehntopf
Founder
Caterina has 10+ years of experience working in business development and operations. She was the second EMEA employee at Uber JUMP, where she launched Berlin operations and defined how markets in Europe operated. Most recently she was the Head of Operations at Dance, an ebike subscription provider, where she built the operations org from the ground up, scaling the company to +10k monthly subscribers with operations in 5 markets.
Caterina Kiehntopf
Founder
Caterina has 10+ years of experience working in business development and operations. She was the second EMEA employee at Uber JUMP, where she launched Berlin operations and defined how markets in Europe operated. Most recently she was the Head of Operations at Dance, an ebike subscription provider, where she built the operations org from the ground up, scaling the company to +10k monthly subscribers with operations in 5 markets.
Mechanical contractors struggle to get timely, accurate updates from the field because reporting is slow and still relies on paper. This leads to project delays, late payments, and frustrated clients.
Dexter’s AI agents make it easy to capture field data using voice, images or documents, enrich it with context, and turn it into structured updates for both construction and service work.
Problem
Completing construction projects on time is notoriously difficult - 75% of construction projects are not done on time - and delays mean financial penalties, strained client relationships, and lost revenue from the next job. The reason for most delays is broken communication between in-field and back office teams.
Field reporting is broken. Many foremen, superintendents and service techs still rely on pen and paper or clunky apps they hate using. This means daily logs and service reports are often late or incomplete.
Project managers are flying blind. Without timely field updates, issues like install errors, weather delays or missing parts go unresolved, delaying schedules and driving up labor costs.
Payments are delayed. Missing reports hold up invoicing and can lead to billing disputes, resulting in cash crunches.
Missed documentation = missed revenue. Change orders, service upsells, and warranty claims fall through the cracks without clear documentation, costing contractors thousands.
Solution
AI agents that work like a new crew member, no training needed.
Voice-powered reporting, built for the field. Dexter’s voice AI lets foremen and techs log updates just by talking - no typing, clunky apps or forms.
Remembers what’s been done, and what’s next. Dexter tracks past progress and follows up automatically, so nothing slips through the cracks on long installs or repeat service visits.
Enriched with knowledge. Integrated with your ERP and project management tools, Dexter pulls in job info like site details, tasks, and deadlines, so the AI starts with context.
Generate structured, client-ready reports. Create customizable reports with photos, equipment, and job details. Upload past examples to easily create templates, so reports are ready to send to GCs, clients, or accounting.
Easily digitize paper documents. Construction sites still rely on paper for deliveries, timesheets, and more. Dexter’s AI can extract all relevant information, verify it against your project plan, and add it to your ERP.
Use Cases
Daily Logs Agent: Dexter’s voice AI calls your foreman or super when they leave the site, asks what got done, automatically pulls in weather and project data, and generates the log for them. The completed log is instantly sent to your PM’s inbox.
Production Reports Agent: Log field hours via GPS tracking, pull material data from your ERP, and compare progress to plan. Your foreman or super uses the voice agent to add context, and the agent turns it all into an actionable report for project managers.
Service Reports Agent: Dexter’s voice agent talks to service techs to collect all required information about the job, including parts used and any follow-ups needed. The agent extracts asset data via photos, and generates a full report for customer approval.
Invoice Match Agent: Dexter digitizes delivery slips, purchase orders, and invoices, flags overcharges or missing items, handles follow-ups with suppliers, and syncs clean data to the accounting system.