AI trained on your business, running where you control it.
Not another rented chatbot. A practical system that uses your orders, policies, documents, catalog, customer history, and internal workflows to do useful work.
People used to buy the CD and play the music whenever they wanted. Now most software asks you to subscribe forever and still accept someone else's rules. DDG's AI services are built around the opposite idea: train or configure the useful intelligence around your data, then choose whether the right path is a Logos fine-tune, a ClawShop vertical package, framework setup, or Erkos-controlled on-prem work.
What decision, reply, forecast, or workflow should improve?
Then we pick
The model, framework, privacy level, and deployment target.
4
Paths
Fine-tune, agent, framework, on-prem
12
Directions
ClawShop roadmap plus Logos examples
0
Model homework
We choose based on workload
1
Buyer question
What outcome should improve?
Which Business Are You?
Twelve directions. One buyer question.
A non-technical buyer should not have to compare Qwen, Llama, or hardware diagrams. They should see their work, their data, the maturity of the offer, and a believable reason this pays for itself.
Local retail, catering, food service
Cookie shop / bakery
FINE-TUNELogos
Demand changes by season, events, weather, and repeat customers. Guessing inventory turns into waste or missed sales.
Uses your
>Historical orders and catering requests
>Ingredient costs and vendor lead times
>Seasonal promos, holidays, and local events
Can produce
>Forecast flour, butter, packaging, and staffing needs before the week starts
>Draft catering replies that match store policy and tone
>Summarize which cookies, bundles, and promos are actually moving
Recommended path
Logos Starter or Studio when the main value is forecasting, policy answers, and reusable store-specific behavior.
This page routes the buyer by the shape of the problem: durable model behavior, productized operator package, framework-only deployment, or regulated on-prem work.
Fine-tune and own the weights
Logos
Train a model on your data and keep the resulting weights.
Best when examples, documents, support history, policies, or brand voice should become reusable model behavior.
Install a business operator agent with productized skills.
Best for stores and planned vertical workflows where the main need is a deployed operator layer: Hermes-Agent bundle, 7 named skills, and Starter/Pro/Enterprise/Custom pricing tiers.
Local models, private frameworks, and on-prem deployments matter because they change the business terms. Your workflow can keep improving around your own data without turning every answer into a metered subscription.
Rented AI
Generic behavior unless you keep prompting
Ongoing usage fees for every interaction
Provider controls the model, policy, and endpoint
Sensitive data may be the limiting factor
DDG setup
Configured around your business examples and rules
Weights, operator package, framework, or on-prem system chosen by need
Local or private deployment paths available
Buyer chooses outcome, DDG chooses the technical route
FAQ
Plain answers for non-technical buyers.
Why would a small business need this?
Not because the owner needs to care about model names. A small business needs it when the same decisions, replies, estimates, orders, policies, or follow-ups happen over and over, and the answers are hidden in that business's own data.
Do I need to know whether I need Qwen, Llama, or another model?
No. We choose the model based on workload, budget, privacy, data shape, and deployment target. The buyer decision is the outcome: what should the system do for the business?
What is the difference between ClawShop and Framework Setup?
ClawShop vertical packages are productized business operators, starting with e-commerce and expanding through the roadmap. Framework Setup is for teams that already know they need OpenClaw, Hermes, NemoClaw, IronClaw, or a custom agent stack installed around an existing workflow.
What if I found a free ClawHub or Hermes skill?
Free skills are useful for technical discovery and proof of the ecosystem. ClawShop is the paid deployment path when you want those skills selected, configured, integrated, documented, and supported for a real business workflow.
When is Erkos required?
Use Erkos when training data is regulated, contractually restricted, air-gapped, or too sensitive to leave your environment. Logos is for non-regulated fine-tuning; ClawShop is for private agent deployment; Erkos is for controlled on-prem work.
Route The Buyer
Bring the problem. We will map the setup.
The first sales conversation should not be a model comparison. It should be a concrete question: what would save time, reduce waste, protect private data, or increase revenue in this business, and which DDG path fits that outcome?