One candidate workflow
A process that is repetitive, rule-based and text-heavy, and costs your team noticeable time each week.
AI agent pilot
Prove it on one real workflow
Before you commit to more, we prove value on one real process, on your hardware, against success criteria we agree with you up front.
Scope
One workflow. Fixed scope. A clear answer.
No multi-year roadmap and no slide decks. Scope and timeline are fixed in writing before any work starts.
- 01 · Scope
Map one workflow
Together we pick one slow or repetitive process, agree the success criteria, and set what the agent may read, do and never touch.
- 02 · Build
Set up on your hardware
We install the platform on your servers, choose the right open AI model and build the agent around your process.
- 03 · Run
Work beside your team
The agent works alongside your people, not instead of them. We measure time saved, error rates and review effort.
- 04 · Review
Results and handoff
You get the working system, its logs, training for your team and a written report on whether to scale it.
Before we start
What a pilot needs from you
A process owner
One person who knows the workflow, can explain the rules and will review what the agent produces.
Access to real examples
Representative documents, tickets or records the agent can learn the process from, inside your network.
Hardware, or a plan for it
Most pilots run on a single workstation with one or two modern NVIDIA GPUs. If you do not have one yet, we help you size it.
Responsibilities
Who does what
Snayu
- Install and configure the platform on your hardware
- Choose and set up an open model that fits the work
- Build the agent around your process and permissions
- Measure results against the criteria agreed up front
- Write the final report and train your team
Your team
- Name a process owner and reviewers
- Provide access to systems and example data
- Provide or approve the hardware
- Review agent output during the run phase
- Decide whether to scale, based on the report
Success criteria
Measured, not promised
We agree the targets with you during scoping, then measure the agent against them on real work.
Time saved
How long the work takes with the agent taking the first pass, compared with today.
Error rate
How often the agent’s output needs correcting, and what kinds of mistakes it makes.
Review effort
How much of your team’s time goes into checking and approving the agent’s work.
Deliverables
What you keep
- The working agent, running on your hardware
- Its logs: every action the agent took during the pilot
- Training for the people who will use and review it
- A written report with measured results and a recommendation on whether to scale
The custom work we build for you is yours, unless your signed agreement says otherwise. The platform underneath, LychD, is open source under MPL-2.0.
Pricing and next steps
After the pilot
The written report gives you a clear recommendation: scale the agent to more of the work, adjust it, or stop. If you scale, the same agent keeps running on your hardware and we agree the next scope in a new statement of work.
Get started
Hand your first workflow to an AI agent, on hardware you own.
Tell us where work stalls, repeats or loses context. We will scope one workflow, run an agent beside your team on your own hardware, and show you in numbers what it is worth.