Consulting
The VM and the Redacy method applied to a team of startup analysts and consultants. The time to complete a review drops from over 30 minutes to 5-10 minutes.
The customer sees a digital employee that becomes specialized in one of their proprietary processes, and the know-how stays theirs.
- Plugin & data organizationIt starts with the data. The free plugin for Claude Code or Codex guides the customer to make the process data accessible, organize it and write the rules of the work. What comes out is ready to go into their AI VM.
- FDE & consultingA company starts with consulting from forward deployed engineers, who work next to the people running the process. On day one they capture the starting point so the return can be measured, and they stay with the company until the agent is at work.
- JVWhen the collaboration works, what was built can become the product of a new company. The partner brings commercial access and knowledge of the problem, Redacy brings the system that scales.
Redacy brings digital employees into the company, specialized in its proprietary processes
Proactive
Processes start on their own, and the agent proposes new rules and tools.
Always learning
It reflects on the work done and distills the company's knowledge from it.
Model agnostic
It works with and adapts to the leading models and harnesses: Claude Code, Codex, dsh, Hermes and others.
Works 24/7
One machine per customer, always on, on-prem or managed by Redacy.
The know-how stays theirs
Data and knowledge stay with the customer. It plugs into the systems they already use: SharePoint, Drive and others.
A command log
Every command leaves a line. Changes go through a person's yes.
We provide the whole AI orchestration layer
From the free plugin that organizes data to FDE consulting that measures the return, from machines working 24/7 to joint ventures for an entire industry: Redacy supports the company or professional at every step.
We've been operating since early 2026. We've sold more than €100k of AI agents, applying the same pattern to different use cases.
Success stories
ESCO
An AI agent for an energy efficiency operator. In 6 weeks it learned to analyze plants and draft efficiency proposals, all the way to commercial proposals. A process that took several people across teams up to 40 hours in total now takes 4 hours of review and approval.
EPC
AI agents for monitoring large photovoltaic plants, at an energy engineering operator. They keep learning new scenarios and respond proactively in the control room. The process takes 90% less time than the previous one, which ran in Excel and by hand.
Acoustic engineers
An agent that works alongside the professional on every activity. After 1 month of learning, it independently handles 35% of the activities that used to take them hours.
The innovation is in the orchestrator
Models, cloud and data come from the market and the customer. The only thing unique to the machine is the system that puts them to work on a proprietary process: orchestrator and CLI.
One agent
It receives every input and routes it to a fixed-step process or a work folder. If there isn't one, it creates it.
Five layers
It builds and maintains the company's AI data, policies, tools, controls and learning.
Earned autonomy
It moves up one action class at a time. An error sends it back down on its own; moving up is a person's call.
Go deeper
Four short pages.
What we're looking for
Redacy is growing with the help of partners, advisors and talented people. Are you one of them?
- Cloud partner
- Hosts customer VMs on a European cloud and handles provisioning. We need one to put the first customers into production.
- FDE
- An engineer with enterprise software experience who stays with the company all the way through. We need one to work with more companies at once. We start project by project.
- Advisors
- Experienced people who help us with open decisions, starting with legal and corporate structure.
- Industry partner
- Someone who knows a vertical and has commercial access. They open the way to trade associations, professional bodies and trade shows, and can become a joint venture partner.
- Distribution partner
- Sells the machines with a Redacy license and invoices the customer. A good way to find customers.
- Data partner
- Handles the sync between the VM and the customer's environment: every customer starts by making the data of their process accessible.
- Harnesses, orchestrators and models
- Claude Code, Codex or others: we integrate the technologies already on the market and stay model agnostic.
- Sensor partners
- Companies specialized in connecting data securely. We need them to build the sensors around the work.
- Investors and programs
- Conversations to spread the idea, starting in November.
Can you help with any of this?