Skip to content

MANAGED COMMUNITIES OF PRACTICE

Expert knowledge that spreads. Expert time that’s protected.

DBX Expert Lab designs and runs managed communities of practice and learning for experts and the teams that rely on them.

  • Today: The same question, over and over.With your Expert Lab: Answer once. Everyone finds it.
  • Today: Crucial issues buried in DMs.With your Expert Lab: They reach your experts, with context.
01 - How it worksProven community habits

Borrow what makes the best communities work.

Merchants, pilots, automation practitioners and peer supporters solve problems together every day in public communities like these. The habits behind them carry over to any field where know-how sits with a few busy experts.

  • Screenshot of the Shopify Community forum
    Shopify CommunityMerchants
  • Screenshot of the let's talk forum
    let's talkPeer support
  • Screenshot of the Infinite Flight forum
    Infinite FlightPilots and players
  • Screenshot of the UiPath Forum forum
    UiPath ForumAutomation practitioners
  • Screenshot of the MongoDB Community Hub forum
    MongoDB Community HubDevelopers
  • Screenshot of the TrustedHousesitters forum
    TrustedHousesittersMembers
  • Screenshot of the Niantic Wayfarer forum
    Niantic WayfarerContributors
  • Screenshot of the YoYoExpert Forums forum
    YoYoExpert ForumsHobbyists
  • Screenshot of the Asana Community forum
    Asana CommunityCustomers
  • Screenshot of the Katalon Community forum
    Katalon CommunityQA professionals
  • Screenshot of the HomeExchange forum
    HomeExchangeTravelling members
  • Screenshot of the Make Community forum
    Make CommunityBuilders
  • Screenshot of the Wyze Community forum
    Wyze CommunityCustomers
  • Screenshot of the Unreal Engine Forum forum
    Unreal Engine ForumGame developers
  • Screenshot of the NVIDIA Developer Forums forum
    NVIDIA Developer ForumsEngineers
  • Screenshot of the Cursor Forum forum
    Cursor ForumDevelopers
What makes them workIn your Expert Lab
Answers stay findable
Answer once. Everyone finds it.Recorded explanationsAn expert explains a recurring problem once, at the lightboard, and the next person who asks finds the answer instead of interrupting them.
Peers answer each other
Answer once. Everyone finds it.Practitioner discussionMembers ask, compare approaches and report what happened in their own work, so experts are not the only ones answering.
Someone notices what keeps coming up
Crucial issues reach your experts, with context.A lead who surfaces the crucial topicsYour community lead tracks the questions that keep returning and brings you the ones that need expert judgment, with the context already gathered.
Fixes are reviewed before they spread
Authorized trials and peer reviewNew methods are tried within agreed limits, then reviewed by peers before other teams adopt them.
Someone keeps it current
Practices with owners and review datesWhat holds up is kept with its reasoning and limits, and checked again when tools, policy or conditions change.
Contributors get credit
Recognized expertsThe people others learn from are named, credited and given time to contribute.

Built-in AI, on your terms

The community does the finding. Your experts keep the judgment.

A dedicated Discourse community brings its own AI. It works from what your experts have already explained, and the community stays useful with it switched off.

Finds the answer

Semantic search and related topics surface the expert's explanation when someone asks in their own words. An assistant can answer from your community's reviewed content and documents, with the source linked.

Protects the lead's time

Summaries catch people up on long discussions. Spam and toxic content are caught automatically. Usage and cost reports show what the AI is actually doing.

Runs where you decide

Discourse's managed hosting with its included open-weights model, your approved cloud AI provider, or your own servers with a locally run model. AI data stays with your community content.

Discourse AI is an open-source capability of the Discourse platform. DBX configures the selected features within your policy; model hosting, local compute and provider fees are scoped separately.

02 - Example · AI adoptionIllustrative example

Give your AI rollout a feedback loop.

People learn where AI helps, where it needs checking and where it should not be used by trying it in their own work. Here is how one finding moves through the community.

Illustrative scenario · Not a client result

The AI draft looks finished. The handover still misses the exception.

Step 1 of 6 · Teams / Slack → community

Surface the problem

“Can we use AI to draft handovers without losing the details the next team needs?” A question in the daily channel opens a focused community discussion.

The DBX community lead invites enthusiastic users, people who stopped, and nonusers. They gather approved or redacted examples, reasons for using or avoiding the tool, and unanswered questions - not just the most confident voices.

Community lead: include quieter voices, follow unanswered questions and moderate respectful challenge.

Possible status at this illustrative stage

  • Proposed (illustrative stage)
  • Trying
  • Reviewed
  • Approved
  • Reused

A question to investigate, not evidence that AI improves the work.

Practice record Illustrative · evidence to collect
Question
Can AI help draft handovers without losing the details the next team needs?
Context
Illustrative cross-team handovers; approved tool and non-sensitive sample material only.
Previous approach
For the trial baseline, document the existing handover and review process before changing it.
Proposed change
Draft from approved sources; check each claim and exception against its source before handover.
Evidence to collect
Reviewed draft/source pairs, omissions, handover quality, drafting time, checking/correction time and a second team's reuse attempt.
Limitations
Sample work only. Preserve failures and exceptions. No inference of production readiness or company-wide adoption.
Reviewer role
Client domain expert and receiving-team peer.
Owner role
Client workflow owner; sponsor and IT own relevant access, tool and policy decisions.
Review timing
Proposed: agree before the trial; revisit after the trial, peer review and the second-team attempt, and whenever the approved AI tool, policy or workflow changes.
Status
Proposed - possible illustrative status only. A question to investigate, not evidence that AI improves the work.
03 - What you getPeople, Station, community

Tools don’t make a community useful. People do.

Every launch brings the people, the recording setup and the community routines together, so your experts explain once and the answers keep working.

An instructor draws on an illuminated lightboard while a nearby monitor displays the recorded explanation.
A senior community lead
One accountable lead runs onboarding, discussion, moderation and follow-up, and brings your experts the topics that need their judgment. The lead is backed by scheduled support in learning design, expert explanation and technical care. Meet the proposed team before you start.
A Knowledge Station
A lightboard with built-in camera and light, set up in your space with operator onboarding. Each explanation comes with captions, a short written companion, a discussion prompt and an application exercise.
The community
It runs in Teams or Slack where they fit, or in a dedicated Discourse community when ownership or deployment needs justify it. Approved policies stay in your systems, and the community stays useful with AI switched off.
Station sizes and setup

35-inch model

Width with legs
39.39 in
Leg footprint depth
11.02 in
Assembly height
22.01-25.95 in
Camera arm
19.33 in

50-inch model

Width with legs
52.47 in
Leg footprint depth
11.02 in
Assembly height
29.37-33.31 in
Camera arm
27.11 in

Published eGlass dimensions, not the full working footprint: the presenter, computer and camera arm need room around the station. You provide a compatible, IT-approved computer and suitable space.

eGlass specifications
Denis Bundi

Who runs your Expert Lab

Denis Bundi

Founder and CEO · Seattle

Denis brings experience in software engineering, instructional design and learning-platform development. He leads your launch with the DBX team.

Denis on LinkedIn
04 - 90-day launchFor organizations

Launch your Expert Lab in 90 days.

A fixed-scope launch, not a subscription. Bring a recurring problem; we set up protected expert time, authorized trials, scheduled reviews and owners who decide.

A presenter writes on a 35-inch eGlass lightboard set on a desk.
35-inch · Desktop

Foundation

Prove the model around one recurring problem.

$15,00090-day launch fee

Practices
One
Initial members
Up to 30
Facilitated reviews
4
Idea launches
2
Complete Lightboard explanations
1
Knowledge Station in the fee
One 35-inch desktop Knowledge Station
Client-controlled deployment
Not included
A 50-inch eGlass lightboard on A-frame legs, with a camera arm extending in front of the glass.
50-inch · Mounting agreed

Enterprise launch

Coordinate related practices across a broader organization.

$35,00090-day launch fee

Practices
Up to 3
Initial members
Up to 150
Facilitated reviews
8
Idea launches
4
Complete Lightboard explanations
3
Knowledge Station in the fee
One 50-inch Knowledge Station, mounting agreed for your space
Client-controlled deployment
Scoped

Station images: eGlass. Shown for reference - final configuration is confirmed in your written proposal.

Every package includes

  • A named DBX community leadWorks inside your organization, in your channels and with your experts.
  • Instructional design supportFor every explanation and exercise.
  • An experience designed with youLearning and community routines built around your experts and your problem, not from a template.
  1. Days 1-30

    Set up the Station, onboard experts and operators, welcome the first members.

  2. Days 31-60

    Record explanations, run discussion and support authorized trials.

  3. Days 61-90

    Review reuse, quality and effort, then decide what comes next.

Good to know

Fees cover the 90-day launch. Ongoing management, hosting, infrastructure, licenses, AI usage, local compute, technical operations and custom integrations are quoted separately unless included in your agreed scope.

  • One station at one client site. Final equipment configuration, delivery charges, tax and any on-site work are confirmed in your written proposal.
  • You provide a sponsor, participating experts with protected time, named decision owners, a compatible computer and suitable space. Unspecified travel and room works are not included.
  • Time, production, platform and client responsibilities are agreed in writing. Complex infrastructure is additional scope.
Plan your Expert Lab
05 - Your own practiceFor independent experts

Running your own practice?

Turn what you know into a cohort course or practice community that builds your standing with the people you want to serve.

Also from DBX: learning delivery and Lightboard Studio recording days.

06 - QuestionsPlainly answered

Questions buyers ask first.

What is a managed community of practice?

A community of practice is a group of people who share a field or a recurring problem and get better at it by learning from each other regularly. The term comes from learning researchers Jean Lave and Etienne Wenger (1991). A managed community of practice adds named responsibility: DBX Expert Lab provides a senior community lead, recorded expert explanations and a review routine, so what experts know is shared, tried in real work and kept current, while your owners decide what changes.

How much does an Expert Lab cost?

Organizational launches are fixed 90-day fees: Foundation $15,000 (one practice, up to 30 initial members), Growth $25,000 (two related practices, up to 75 initial members) and Enterprise launch $35,000 (up to 3 practices, up to 150 initial members). Each includes one Knowledge Station at one client site. Every package includes a named DBX community lead working inside your organization, in your channels and with your experts; instructional design support for every explanation and exercise; and a learning and community experience designed with you, not from a template. Ongoing management, hosting and licenses are quoted separately. For independent experts, course design starts at a $750 course base plus $225-$1,125 per lesson.

How is a community of practice different from training or an online community?

Training teaches a set curriculum, usually once. An online community hosts conversation, but often without anyone responsible for what it produces. A community of practice is organized around a shared, recurring problem: members try methods in their own work, compare what happened and keep what holds up. An Expert Lab combines recorded expert explanations, practitioner discussion, authorized trials with peer review, and practices that have named owners and review dates.

How long does it take to launch a community of practice?

An Expert Lab launch takes 90 days, in three phases. Days 1-30: Set up the Station, onboard experts and operators, welcome the first members. Days 31-60: Record explanations, run discussion and support authorized trials. Days 61-90: Review reuse, quality and effort, then decide what comes next. After 90 days you can extend, take a handover to run it yourselves, or stop. Ongoing management is quoted separately.

Why pay for this when we already have Microsoft or Slack?

You already own capable collaboration tools. DBX's fee pays for named responsibility: senior community leadership, expert participation, learning design, facilitation, maintained resources and follow-through. If your existing tools fit, we work there. A separate platform such as Discourse is recommended only when durable discussion, client-controlled deployment, portability or ownership justify it.

Where does it run, and do we have to use AI?

AI is optional, permissioned and reviewed, and the community is designed to be valuable with it switched off. The community can run in your existing tools, on managed Discourse hosting, or on your own infrastructure, on premises or in a private cloud, within an agreed scope. A dedicated Discourse community includes Discourse AI: semantic search, related topics, summaries, assistants that answer from your reviewed content, spam and toxicity detection, and usage reports, with AI data stored alongside your community content. Discourse's included hosted model comes only with its hosting plans; self-hosted communities connect approved cloud AI, where selected content is processed by that provider, or separately scoped and validated local AI. Identity, access, retention and the processing boundaries for AI, email, analytics and backups are agreed with your IT team.

What if we already have a thriving community?

Keep what is working. We target agreed gaps, such as expert workload, learning design, cross-team reuse or maintained practice, rather than replacing a healthy community.

How will we know it is working?

We record a baseline at the start, then look for reuse beyond the originating team, quality of methods, changes in expert effort and value to members. No posting quotas, and no outcome is guaranteed.

What is included with the Knowledge Station?

The package's selected eGlass lightboard with built-in camera and lighting, agreed mounting, setup and operator onboarding at one client site. You provide a compatible, IT-approved computer and suitable space; if you already have recording equipment, we reuse what is suitable. Foundation includes one complete explanation, Growth two and Enterprise launch three, with lengths and revisions agreed in the proposal. Delivery charges, tax, audio, cabling and any on-site work are confirmed there too. Unspecified travel and room works are not included.

Plan your Expert Lab

Bring us the questions people keep asking you.

In your summary, tell us:

  1. The questions or problems that keep coming back to you or your experts.
  2. Who needs the answers: your teams, your clients or your audience.
  3. Where they already gather, such as Teams, Slack or an existing community.

DBX Expert Lab
Denis Bundi, Member-Manager & CEO
Seattle-based · Working nationally
denis@dbxai.com
206-880-2045

Where does your expertise live?

Working independently? Enter your practice name or your own name.

What happens next

DBX reviews your initiative. We’ll clarify the audience and challenge, then discuss whether DBX is the right fit for a scoped next step. A brief description is enough to start.

Prefer email? Contact denis@dbxai.com directly.

How we use your information

DBX saves the details you submit to review your initiative, respond, and follow up. We also record the DBX page you came from and the progress of the conversation to understand which inquiries are a good fit. Submitting does not sign you up for a newsletter.

Please do not include confidential client or personal information in your description. For questions about your submitted information, or to request its deletion, email denis@dbxai.com.