Decision AI
Decision AI — a voice-first AI simulation platform helping managers practice high-stakes business decisions under pressure.
Role
Product Designer
Year
2026
Platform
Web — Desktop & Mobile
The Problem
Managers are expected to make important decisions quickly, often with incomplete data, competing priorities, and stakeholder pressure. Workplaces are moving faster, and there is less time to analyse, align, and decide.
Companies need employees who can interpret business data quickly, defend their reasoning, and handle pushback from senior stakeholders — while balancing short-term and long-term consequences.
Today, most decision-making capability is built through real-world experience. That is slow, inconsistent, and risky.
Learners rarely get a safe way to practice business judgment, make mistakes, see consequences, and improve before the stakes are real.
The Concept
A safe room to practice the decisions that actually carry consequences.
Decision AI is a voice-first simulation, not a chatbot. Learners enter a realistic corporate scenario, review case data, and hold a live voice conversation with an AI manager or stakeholder who challenges, questions, and pressures them the way a real executive would.
Every decision branches into a consequence, and every session ends in a structured debrief on judgment, reasoning, trade-offs, and decision quality — built for enterprise clients training mid-level and senior managers.
Decision AI vs. Convo AI
Our existing Convo AI product trains the behavioral layer — how to handle a conversation. Decision AI trains the business layer — how to read data, weigh trade-offs, push back, and commit to a defensible call with a senior stakeholder.
| Dimension | Convo AI | Decision AI |
|---|---|---|
| Training Layer | Behavioral — conversation quality | Business — judgment, reasoning, trade-offs |
| AI Behavior | Responds to learner dialogue | Challenges, questions, and pressures the learner |
| Feedback | Mostly behavioral | Decision quality, business impact, risks, consequences |
| Conversation Flow | Continues smoothly | Choices create branching consequences |
| Best For | Soft skills | Business, leadership, strategy, risk, management capability |
Target Users
Mid-Level Managers
Frequently make team, client, operational, or business decisions. Own execution and short-to-medium-term outcomes, need to balance speed with accuracy, and often face stakeholder pressure. May struggle with ambiguity or incomplete data, and need practice defending recommendations as they grow into stronger leadership roles.
Senior Managers
Make higher-impact decisions across teams, clients, or business units. Comfortable with complexity but still need sharper judgment on decisions involving revenue, risk, clients, and strategy. Need to influence stakeholders while staying accountable, and expect a premium, credible enterprise experience.
What Learners Need
Learners need to
- →Understand the business scenario quickly
- →Review relevant data without feeling overwhelmed
- →Practice making decisions under realistic pressure
- →Be challenged by an AI manager or stakeholder
- →See the consequences of their choices
- →Learn what they did well and what they should improve
- →Leave with clear action items
The Learner Journey
Each stage carries a distinct emotional posture, and the design escalates that posture from screen to screen.
Briefing
Curious, settling in. The design feeling is cinematic and inviting — a moment to orient before the pressure begins.
Caselet
Focused, absorbing. Editorial, dense-but-clean — data is presented like a dossier, not a dashboard.
Gameplay
Engaged, slightly pressured. Charged, present-tense, in-the-room — the learner feels like they're actually in the meeting.
Debrief
Reflective, vulnerable, hungry for clarity. Constructive, structured, premium-report — consequences resolve into insight.
Gameplay — Inside the Simulation
The gameplay screen is a three-panel layout, each panel carrying a different weight and a different job.
Left — Caselet & Data
Always available but visually recessive relative to the center — lower contrast, smaller scale, a 'paper' feel. It's the dossier on the table, not the conversation. Collapsible and expandable so a learner can pull data forward when needed, with sticky tabs across multiple data views.
Center — Conversation
The visual hero — where the learner spends 80% of their attention. AI turns are named and role-tagged ('Priya Menon, CFO') and read with authority. Three switchable input modes: voice (primary), text, and AI-offered multiple-choice. Voice state — idle, listening, transcribing, AI thinking, AI speaking — is always visually obvious. MCQ convergence moments are treated with the gravity of a submit, not a casual tap.
Right — Meeting Room
Visualises the AI manager, stakeholders, and the learner, and communicates who is speaking right now. Stylised avatars and abstracted presence cards over literal video-call mimicry, with subtle ambient motion when the AI is thinking or speaking. This panel earns its place by reinforcing 'I am in a meeting, not in a chatbot.'
Desktop
Showing slide 1 of 7: Briefing
Briefing
Mobile
Showing slides 1–3 of 7
Debrief
The debrief resolves the session's tension. Where gameplay is charged and present-tense, the debrief is reflective and structured — a premium report on performance, cognitive insight, session playback, and strategic next steps.
Desktop
Showing slide 1 of 6: Simulation Summary
Simulation Summary
Mobile
Showing slides 1–3 of 5
Defensible Position
Judgment, Not Just Conversation
01
Judgment Over Dialogue
Trains business reasoning and trade-offs, not conversational fluency. The AI challenges and pressures the learner rather than simply responding to them.
02
Consequential, Not Conversational
Choices branch into real outcomes. Where Convo AI lets a conversation continue smoothly, Decision AI doesn't let you off easy — every call has weight.
03
Enterprise-Grade Presence
A premium, credible experience built for senior-manager stakes — composed and consequential, distinct in personality from Convo AI's warmer tone.
Same brand family as Convo AI, distinct product personality. If Convo AI is warm and conversational, Decision AI is composed and consequential.
Visual Direction
Boardroom, not bootcamp — sophisticated and slightly serious, with moments of energy. An exploratory direction, designed to feel composed and consequential, distinct from Convo AI's warmer, more conversational personality.