Mail · chat · meetings · calendar
Inbox, threaded chat, video meet, shared calendar — one identity, one search across the lot
Experience and efficiency The distance still left to close
Outside, turn your website into an AI-native interface. Inside, let AI staff join your organisation.
Products
Two products
live 
dddk SDK
Published on npm, running in production. Analytics dashboard + small-model training are the next layer.
- shippedwebagent · palette · voice · dwell · 4 interaction surfaces
- nextAnalytics dashboard · small-model lab · cohort insight
- horizonEdge ASR / TTS · non-voice signal modalities
wip 
Enterprise platform
Built for human-and-AI collaboration. AI staff each have their own identity, scoped permissions, audit trail — same surface as the rest of the team. Data, keys, and models stay in your hands.
- in buildComms · collab · ops · HR · ERP · CRM · Admin modules
- releaseDate not yet announced
- AI staffIdentity · permissions · audit trail · reply style
dddk side
From experience, weaving stability, models, and growth into three paths
Every command, every voice clip, every page-understanding pass becomes a signal
From the dashboard, the stream forks three ways — audit, models, marketing
- 01 · dddk collects
- SDK installed; every action is a signal — hot features, dragged paths.
- 02 · Data analytics
- Dashboard reveals what your users actually do. The stream forks three ways from here.
- 03 · Audit logs
- The same signal stream as stability view — catch regressions and slow paths before users complain.
- 04 · Small-model training
- Use accumulated domain data to train a customer-specific small model — more accurate, cheaper, controllable than a general LLM.
- 05 · Marketing insight
- Find the highest-value cohorts and which features deserve the spotlight.
Enterprise side
Enterprise management platform
Mail, docs, tasks, HR, ERP, CRM — every kind of work the org needs, on one surface. No ten-SaaS subscription stack to switch between.
AI staff use the same surface as the team: same identity, scoped permissions, audit trail
Docs · Wiki · Drive · workspaces
Real-time docs, shared wiki, files, scoped workspaces — on your own data plane
Tasks · projects · workflows
Per-workspace task boards, statuses, workflows, project closeout — AI staff can be assigned tasks and step into the flow
Onboarding · leave · shifts · benefits · offboarding
Members, departments, leave, shifts, benefits, devices, offboarding — the entire employee lifecycle, one surface
Accounting · purchasing · expenses · assets
Bank import, journal, purchase orders, fixed assets, budgets, expense review — finance and assets on one trail
Contacts · companies · activities
People and company records, activity history, mail tie-in — the layer where the org meets the outside world
Admin
Permissions · approvals · audit · security
The spine the other modules sit on. Field-level permissions, immutable audit trails, approval workflows — every employee, human or AI, is bound by the same rules. The backbone that keeps an AI-heavy org stable at scale.
AI staff
AI staff are colleagues, not tools
Each AI staffer has its own account, scoped permissions, audit trail. @mention them, assign them tasks, invite them to meetings, ask them to draft a contract. Manage them like employees. As more of the org runs on AI staff, the permissions + audit layer above is what keeps scale from going feral.
Commercial
Three ways to work together
Embed dddk into your product · roll Enterprise out across your organisation · co-build a SaaS line with us
From a one-off licence to a multi-year revenue share
dddk Commercial License
Use dddk in a closed-source product, run a SaaS, or embed it in an external service — the commercial license waives the AGPL-3.0 source-disclosure clause. Annual fee, smallest tier is free.
- Under US$500k revenue · Free
- US$500k – US$2M · US$3,000 / yr
- Over US$2M · From US$15k / yr
Custom Development
We build the product for you. From dddk integration, prompt engineering, and agent flow design, through to general software features. Project-based pricing, billed against progress.
- Product / SDK custom builds
- Existing-system integration
- AI workflow / agent design
SaaS Co-op
We build a complete SaaS to live operation, then hand it to the right operator. We keep shipping AI + product upgrades; the partner grows the market. Exclusive license per market, upfront fee + long-term revenue share. The revenue share IS the price of exclusivity — if it ends, we re-license to another operator in the same market.
- Exclusive market / industry license
- Upfront fee + long-term revenue share
- Ongoing AI + product upgrades