MockMarket

Building is easy. Building something that people will actually use is hard. Stop guessing, start validating, and stress-test your ideas against the reality of the vast market in minutes.

Comment

MockMarket

Building is easy. Building something that people will actually use is hard. Stop guessing, start validating, and stress-test your ideas against the reality of the vast market in minutes.

The Problem

Most startups and campaigns fail because they build things nobody actually wants. The traditional feedback loop is fundamentally broken:

  • The Echo Chamber Trap: Founders pitch to friends, family, and colleagues who are too polite to tell them their go-to-market strategy is flawed or their unit economics do not make sense.
  • The Access Barrier: It is exceptionally difficult to secure time with your exact target customers, especially high-profile enterprise buyers or Tier-1 investors. You rarely get in the room with them until it is time to sell or raise, and by then, a fatal flaw means the bridge is burned. MockMarket solves this by simulating these exact stakeholders. Because our agents process live market research before the meeting, they accurately reflect the real intelligence, needs, and pain points of the people you must convince.
  • Wasted Runway: Founders waste months, and thousands of dollars, building products that already exist or solving problems no one will pay for, simply because they could not validate their idea under real pressure.

The Solution

MockMarket is a concurrent, multi-agent simulation engine that replaces weeks of user research and due diligence with a 5-minute, real-time stress test.

By leveraging the Gemini Live API, we move far beyond the text box. Users step into a live audio "Hot Seat" and pitch directly to a panel of autonomous, hyper-critical AI personas. The system orchestrates real-time conflict, analyzes your multimodal inputs, and translates the chaos into consulting-grade, actionable data.

Limitless Multimodal Use Cases

Because MockMarket allows you to define hyper-specific custom agents and stream your screen or camera directly to their vision models, the platform acts as an infinite, on-demand testing sandbox.

Product and Strategy

  • A/B Testing on the Fly: Share your screen to show two different UI designs side by side. Ask a custom board of "Target Users" which one solves their pain points better.
  • Pricing Validation: Pitch your subscription model to a "Frugal CFO" agent to see if your unit economics justify their budget.
  • Feature Prioritization: Talk through your roadmap and let the AI debate what is a "must-have" versus a "nice-to-have" before you write a single line of code.

Brand and Marketing

  • Campaign Simulation: Share your screen to play a new video advertisement. Ask a panel of custom demographic personas for their raw, emotional response to the branding.
  • Message Testing: Read your website copy out loud and see if the AI actually understands your value proposition, or if they find it confusing and saturated.

Public Sector and Advocacy

  • Crisis Management: Define a board consisting of a "PR Critic", a "Liability Counsel", and an "Angry Voter". Read your draft press release and simulate the public backlash before it goes live.
  • Policy Framing: Pressure-test new public sector announcements against opposing viewpoints to discover vulnerabilities in your argument.

Core Technology and Features

  1. Live Market Intelligence: You do not just pick generic bots. MockMarket uses gemini-3.1-flash-lite-preview to instantly scrape the web for your actual, real-world competitors. It then dynamically synthesizes custom agents tailored specifically to attack your unique market vulnerabilities.
  2. The Multimodal Hot Seat: Talk is cheap. Turn on your camera to show the board a physical hardware prototype, or share your screen to walk them through your financial models. The AI evaluates your visual evidence in real-time.
  3. Native Real-Time Voice: Experience zero-latency, realistic voice conversations powered directly by gemini-2.5-flash-native-audio-preview. This is not a text-to-speech robot reading a script; it understands pacing, tone, and inflection, allowing for natural interruptions.
  4. Concurrent Orchestration: This is not a chatbot wrapper. We parallelized agent reasoning so personas think and react simultaneously. A strict persona management layer prevents "drift", ensuring a skeptical agent remains fiercely skeptical no matter what you say.
  5. Actionable Reports: We bridge the gap between creative AI roleplay and rigid statistical analysis. When the session ends, the engine parses the messy, unstructured audio transcript and generates a highly structured PDF playbook detailing exact pivot trajectories, risk levels, and a timeline of exactly when you lost the room.

Technical Architecture

MockMarket relies on a bifurcated architecture to handle complex WebSockets, real-time audio streaming, and frontend rendering seamlessly:

  • Frontend: Built in Flutter (Dart) for a highly responsive, cross-platform UI, hosted live on Firebase Hosting.
  • Database and Auth: Firebase Authentication and Firestore securely save and manage the generated Autopsy Reports.
  • Backend Bridge: A custom Python (FastAPI) server utilizing WebSockets. This backend acts as the crucial bridge for the Gemini Live API, streaming raw PCM audio bytes and base64 video frames between the browser and Google's servers with near-zero latency.
  • AI Models Used:
    • gemini-2.5-flash-native-audio-preview: Used exclusively for real-time, low-latency voice streaming and vision analysis in the Hot Seat.
    • gemini-3.1-flash-lite-preview: Powers all other intelligent features, including heavy JSON logic parsing, parallel orchestration scoring, Autopsy Report generation, and rapid pre-pitch market research.

Built With

Share this project:

Updates