Join a session

Got a code? Hop in.

Enter the join code your facilitator shared.

What is DecisionBuddy?
Live AI facilitation · Real time, in the room

The meeting you wish you'd been running for the last five years.

DecisionBuddy is the AI co-facilitator that joins your workshop in real time — tracking what's covered, surfacing what's missing, parking what's tangential, and turning the whole conversation into a decision-ready artifact before anyone leaves the room.

See how it works
◆ Auto-checks tracker items as you discuss them
◆ Flags one-sided arguments before you decide
◆ Wrap-up artifact ready in 90 seconds
What it actually is

Decision infrastructure for the room you're already in.

A privacy-focused AI workspace for structured discussions, accountable decisions, and reusable organizational knowledge. DecisionBuddy helps groups think clearly, decide responsibly, and preserve the reasoning behind each decision — so the artifact you walk out with is the one your organization actually needs.

◆

Structured discussion

Phase-aware AI that knows when to listen, when to challenge, and when to converge. Not a transcription tool with a chat bolted on.

★

Accountable decisions

Every decision is captured with the reasoning, the dissent, and the people behind it — not just the outcome.

↺

Reusable knowledge

Wrap-ups become artifacts your organization can actually re-open a quarter later. Decisions, sources, follow-ups — searchable.

We are NOT a mere replacement DecisionBuddy doesn't replace Teams, Zoom, or your note-taker. It runs alongside them and produces the one artifact those tools never give you: a defensible record of how the group decided.
What it actually does

It's not a transcript.
It's another participant.

Most tools listen and write things down. DecisionBuddy listens, thinks, and helps the group think better — surfacing the right card at the right moment.

01
Trackers auto-check. The AI ticks off discussion items as the group covers them — and flags the ones being skipped.
02
Gaps get surfaced. If the group's only discussing benefits, the AI gently asks about risks. Devil's advocacy without the awkwardness.
03
Tangents get parked. Brilliant side-thoughts don't derail the conversation — they land in a structured parking lot for review.
04
The facilitator stays in charge. Every AI suggestion is approve / dismiss / snooze. Never auto-acts. Never interrupts.
Event Log · AI Tooling — Q2 Deep Dive
LIVE · 23:14
Mistral
Privacy
Integration
Risks
The wow feature

The tracker that auto-completes itself.

When a topic begins, the AI proposes a tracker — pricing, privacy, integration, risks. As the group discusses, items tick off automatically. If the conversation moves on with items unchecked, you get a card: "You haven't discussed Security yet."

  • Facilitator-approved. AI-suggested trackers wait for your tap before reaching the shared log.
  • Skip detection. Unchecked items aren't quietly forgotten — they're flagged with an amber mark and routed into the wrap-up.
  • Re-park & revisit. Realised something needs more work? Send it back to the parking lot in one click.
  • Resolution tags. Each checked item is tagged: discussed, decided, deferred, assigned.

Topic: Mistral API integration

Decision phase
✓
Pricing & quota model
✓
Privacy & data residency
·
Integration approach Discussing now
!
Performance benchmarks Skipped
Rollback plan
Owner & timeline
!
Heads up — the group is moving on, but performance benchmarks were never discussed. Park, schedule, or address now?
Grounded in facilitation science

An AI that knows when to be quiet.

Every workshop cycles between divergent exploration and convergent decision-making. DecisionBuddy has six explicit phases — and adapts how it shows up in each one. Capturing in Discovery. Assertive in Decision. Synthesising in Review.

  • Discovery phase: Trackers and gap alerts suppressed. Just listen, cluster, and capture.
  • Decision phase: Trackers, completeness checks, transition cards — full assertiveness.
  • You change phases. AI suggests transitions — never forces one.
  • Three assistance levels. Quiet / Balanced / Active — set per session.
Assertive & structured

Decision

Narrowing options and making choices. The AI runs trackers, auto-checks coverage, surfaces hidden assumptions, and offers transition cards before you move on. This is where it earns its keep.

Devil's advocate, structured

The voice that asks "but what about…" so nobody has to.

Every group has a tendency to converge too fast. DecisionBuddy watches not just what you discuss — but how one-sided the discussion is. When perspectives are missing, it surfaces them. Quietly. To the facilitator. Never to the room.

  • One-sided arguments: "You've discussed benefits but no risks."
  • Vague agreements: "The group agreed but didn't say who, when, or how."
  • Hidden assumptions: "This decision assumes X. Has the group validated that?"
  • Circular discussion: "This point has been revisited three times. Decide, park, or table?"
!
Perspective gap · 18:42 · facilitator only
One side of the argument is missing
The group has discussed benefits of Mistral for 11 minutes but hasn't raised any risks, costs, or alternatives. Consider surfacing a counter-perspective.
◆
Vague agreement · 18:45
"We'll handle it" needs a name
The group agreed on migrating the auth layer, but no owner, deadline, or success criteria was named.
◐
Hidden assumption · 18:51
Worth checking before you commit
This plan assumes EU data residency is non-negotiable. Has the group validated this with legal?
The full toolkit

Everything else a workshop needs
— without the friction of bolting it on.

Built into one event-log-centric interface that participants barely notice and facilitators can't imagine working without.

⏱

Catch Me Up

Joined late? Stepped out? Tap once — get a 3-sentence personalised briefing of what's been decided, parked, and asked since you last engaged. Under 3 seconds.

P

Parking Lot

Park ideas mid-conversation with a tap, voice snippet, or transcript bookmark. Each park gets notes, files, and a clear destination — promote to topic, add to tracker, or carry forward.

★

Live Highlights

The AI surfaces concrete, positive recognition in real time — "James brought pricing data that grounded the discussion." Praises only. Never shames.

⊞

Participation Equity

Tracks signal usage, chat, and research per participant. Surfaces imbalance to the facilitator: "3 people haven't weighed in." No callouts to the room.

⚡

Disagreement Sensing

Detects unresolved tension — repeated objections, terminology mismatches, dominance patterns. Surfaces it gently to the facilitator, never auto-resolves.

◉

Decision Readiness

A live readiness bar shows the criteria for the active topic — alternatives compared, owner identified, risks aired. Forward momentum without the leaderboard.

⏲

Time-Aware Facilitation

Pace checks, hard-stop reminders, phase budget drift. "Topic 1 took 25 minutes of 60. Later topics may get squeezed." Never interrupts. Never guilt-trips.

◇

Engagement Styles

Participants self-select on join: Discusser, Researcher, Synthesizer, or Challenger. Adapts the UI defaults — never locks anyone out.

⌕

Research Mode

Anyone can run a private lookup mid-meeting — pricing, competitor docs, fact-checks. Share to the event log when it's worth everyone's attention.

▣

Knowledge Base + History

Files, transcripts, prior-meeting decisions, open action items, carry-forwards. Searchable. Source-badged. Cross-session memory baked in.

↯

Phase & Flow Suggestions

Steel-manning when deadlocked. Round-robin when one voice dominates. Dot voting when there are too many options. A toolbox of techniques offered, never imposed.

▼

Decision Votes

Trigger a formal vote on any proposal. Yes / no / abstain tally with thresholds and pass/fail. Decisions stop being implicit — they become explicit, transparent, archived.

~10s
Live transcription latency
90s
Wrap-up artifact, finalised
6 phases
Phase-aware AI behaviour
0 lock-in
LAN-aware · BYO inference
The other half of the product

The wrap-up isn't an afterthought.
It's half the value.

Every workshop you've ever attended ended in a vague sense of "we agreed on something." A 5-phase wrap-up replaces that with a finalised, collaboratively-reviewed artifact — in 90 seconds.

Pre · Draft · Feedback · Regen · Final

AI Tooling — Q2 Deep Dive

Thu Apr 23 · 6 participants · 47 min

Impact summary

Decisions
4
Action items
7
Parked items
3
Topics covered
5

A 90-second window for participants to correct the AI's first draft. Star ratings, one-word reactions, additions, deletions — all incorporated into the regeneration. Collaborative correction, not passive consumption.

Decisions
4
Adopt Mistral for transcript redaction · 6/6 yes
EU-only data residency required · 5/6 yes
Defer benchmark plan to next session · deferred
Action items
7
Draft data-flow diagram · Ash · Apr 30
Get legal sign-off on residency · Tobias · May 2
Compare costs vs. self-hosted · Lucas · May 5
Follow-up suggested
1
"Mistral rollout — benchmark & rollback plan"
Suggested attendees: Ash, Lucas, Jessica · 30 min · agenda drafted
Built for the whole room

One session. Three views.
Everyone gets exactly what they need.

Asymmetric by design — because what a facilitator needs to see is fundamentally different from what a participant needs, which is different again from a stakeholder watching from outside.

Facilitator

Cockpit view

Three-panel command panel. Topic timeline, phase dropdown, AI-pause toggle, equity observations, transition cards. Coverage at a glance.

Three-panel layout, drag-and-drop tabs
Phase & assistance level controls
Tracker, parking, knowledge, people, PoCs
Decision readiness bar below timeline
Facilitator-only equity observations
Mistral
Privacy
Integration
▸ tracker · ▸ chat · ▸ research · ▸ parking · ▸ kb
$ phase = decision · level = balanced
Participant

Companion view

Two-panel layout. Event log left, sidebar right. Signal buttons, private and public chat, personal research. Mobile-first overlay.

Zero-friction QR-code join
Agree / Disagree / Go Deeper / Want to Speak
Private chat + Public group chat
Personal research (sharable)
Engagement style: Discusser / Challenger / …
▸ event log · ▸ transcript
[ Agree ] [ Disagree ] [ Deeper ]
Want to speak ↑ active
Observer

Read-only view

For stakeholders who can't be in the room but need to follow along. Event log + transcript. No noise, no inputs.

Two-panel read-only layout
No signals, chat, or research
Live event log with phase markers
"Observer" badge in header
Mobile: transcript stacks below
▸ event log · ▸ transcript
[ inputs disabled ]
badge: observer
Where it fits

Built for workshops.
Useful anywhere a group has to decide something.

The mechanics — coverage tracking, gap detection, decision capture, structured wrap-up — apply far beyond the design-thinking workshop. A few of the rooms we already see DecisionBuddy in.

◆

Workshops

Strategy days, design sprints, kickoffs, retros. The original use case — coverage trackers, parking lot, decision votes, collaborative wrap-up. Where it all started.

Strategy · Design · Retro
⏱

Daily team meetings

Standups, project syncs, planning. "Catch me up" for anyone who joined late, parking lot for off-topic threads, and a 90-second wrap-up that becomes the meeting notes nobody has to write.

Standup · Sync · Planning
⚖

Council & city-hall meetings

Local government discussions where data privacy is non-negotiable and the artifact has to stand up to scrutiny. Runs on your network, with your inference. Speaks the vocabulary you already use.

Stadt · Lenkungskreis · Arbeitskreis
✓

Compliance & policy reviews

Decisions that need a defensible paper trail. Every choice ships with the reasoning, dissent, sources, and timestamp — searchable months later when someone asks "why did we agree to that?"

Audit · Risk · Policy
Pricing & service

Software alone won't change a meeting. Ours doesn't ship without us.

We want you to get a real return — not a tool that gathers dust. That's why DecisionBuddy never ships as bare software. Every license comes packaged with the setup, consulting, integration, and change-management work it takes to make decisions actually land in your organization. A single strategy workshop already costs you more than this in salaries, prep, and follow-up writing — we've just made the wrap-up appear before the room clears.

Standard
€23,500 Setup + first year included

Everything one pilot team needs to make decisions stop slipping — from staged hardware to first wrap-up in inboxes.

  • Included
  • White-label setup — your logo, your colors, your terminology baked into the room view
  • Hardware install — laptops, mics, projector, network, all staged before the room opens
  • Environment configured on your infrastructure — LLM endpoints, prompts, guardrails
  • Knowledge integration — connect one system: Confluence, Notion, SharePoint, or Drive
  • Initial change management — kickoff workshop, facilitator training, rollout playbook
  • First year of subscription — included, no surprises
  • Versioned prompts — you choose when the model changes underneath you
  • Year 2+ pricing — usage-based, scaled to the rooms you actually run
Start a pilot
Most teams choose this
Advanced
€64,800 Setup + first year, org-wide

For when one room becomes the whole organization. SSO, deeper integrations, and a change programme to make it stick.

  • Everything in Standard, plus
  • SSO via SAML or OIDC — Okta, Azure AD, Google Workspace — one click in
  • Multiple knowledge integrations — wire DecisionBuddy into the systems you actually use today
  • Custom decision-artifact templates — match your existing reporting and governance standards
  • Extensive change-management programme — train-the-trainer, champion network, adoption metrics
  • Local-first / on-prem inference — your network, your inference, your data, fully air-gappable
  • Dedicated success manager — one named human, not a ticket queue
  • Quarterly model & prompt review — we tune as your domain evolves
  • Priority support — same-business-day response, named escalation path
Talk to us
Enterprise
Custom Tailored to your organization

Multi-region, regulated, sovereign. Co-built with your security, legal, and procurement teams from day one.

  • Everything in Advanced, plus
  • Sovereign / air-gapped deployment — fully isolated, your hardware, your network
  • Multi-region & multi-language facilitation — localized prompts, localized rooms
  • Custom AI development — domain-specific models, fine-tuned prompts for your work
  • Co-developed roadmap — your priorities shape what ships next
  • Compliance support — GDPR, ISO 27001, sector-specific (finance, healthcare, government)
  • 24/7 support with SLAs — incidents handled around the clock
  • Unlimited business units — one programme covers the whole organization
  • Dedicated infrastructure — your shard, isolated from other customers
Request a quote

How the money works: the upfront fee covers setup, consulting, white-labeling, integration with your knowledge systems, change management, and your first year of subscription. From year two, you pay a usage-based subscription scaled to the rooms you actually run — no minimums, no shelfware.

Every workshop you've run since 2020 was missing this exact tool.

Run one session with DecisionBuddy. Watch a tracker auto-complete itself. Watch a tangent get parked instead of derailing the room. Watch the wrap-up appear, edited collaboratively, in 90 seconds. Then try going back.

See a sample wrap-up
✓ Runs on your local network
✓ Works on phones, tablets, laptops
✓ Zero participant install
✓ Bring your own LLM
NeuroForge · AI workshop facilitation

Curious what an AI decision buddy can do?

DecisionBuddy is a private engagement. Reach out to NeuroForge for a live walkthrough, a pilot, or to talk about whether it fits your team.

Contact NeuroForge See a sample wrap-up
or write to kontakt@neuroforge.de