STATUS: Research Prototype (Complete)MODE: Open Evaluation

Public Demo

This demo is openly accessible for evaluation and exploration. Usage is intended for short sessions and architectural review, not prolonged production use.

EXPERT CLUB AI

Agentic AI Workflow Engine · Cognitive Conflict Simulator

Next.js App Router (BFF)React 19 (Streaming UI)Firebase Firestore (NoSQL)Firebase AuthOpenAI / DeepSeek / GeminiServerless Edge Runtimes
[SYSTEM_OVERVIEW]

⚠️ PROJECT STATUS: Advanced Architectural Showcase. Engineered strictly as a live technical sandbox for engineering teams and CTOs to evaluate my multi-agent orchestration skills.

This is a purpose-built demonstration of LLM reliability, streaming UX, and complex cognitive protocols. Deployed specifically for technical review and stress-testing under the hood, not as a commercial SaaS.

The system addresses a structural weakness of single-assistant AI tools: confirmation bias and shallow validation. Instead of reinforcing user assumptions, it intentionally introduces structured intellectual conflict.

The core idea is not 'better answers', but better thinking under pressure. It simulates multiple conflicting expert perspectives, enforces cognitive discipline through protocolized reasoning, and orchestrates debates as repeatable computational processes.

Core Loops
Configure ExpertsInject ContextCompile Persona (System Prompt)Stream Debate (Turn-based)Parse Reasoning Trace vs SpeechJudge Analysis

/// DEPLOYMENT_SCOPE

  • ::Founders & Strategists
  • ::Deep Reasoning Engine
  • ::Not a Chat Wrapper
  • ::Repeatable Debate Protocols
  • ::Feature-Complete for Current Scope

Engineering Architecture

Monolithic BFF (Next.js)

Intentionally avoids microservices to keep iteration speed high and agent logic co-located. Business logic lives inside API routes for maximum velocity at the current iteration stage.

Stateless Agent Architecture

An Agent is not a running process but a data-driven contract defined by 'ConfiguredExpert' (Identity, Cognitive Structure, Domain Bias). It exists only for one HTTP request.

Dirty Realism (Adversarial Protocol)

Anti-fluff, adversarial critique, structured output. Models are trusted for content, not structure. A custom state machine parses streams in real-time.

Sliding Window Memory

Context is reconstructed every turn. History is truncated (MAX_MSGS=30) to prevent runaway token costs and hallucination amplification.

AI Engine

Dynamic Expert Compiler

Instead of static system prompts, the backend compiles a unique persona for every turn based on the expert's archetype (Analyst/Synthesizer/Resonator) and current debate state.

Judge Meta-Agent

A separate /api/judge meta-agent evaluates the full discussion thread to produce structured analysis, remaining independent of the debate participants.

Stream Parsing Engine

Custom stream parsing across backend + UI separates reasoning trace ([THOUGHTS]) from public speech, allowing the user to see the process without waiting for full completion.

Runtime Persona Constructor

Parametric Definition (Scalars -> Vectors)

Users do not write prompts. They manipulate graded scales (Archetype Mix 0-100%, Traits 1-10 like 'Conformism' or 'Openness'). These parameters are stored as structured JSON, not text.

Heuristic Bucket Compiler

A runtime compiler translates numerical values into nuanced behavioral instructions using non-linear heuristic buckets. (e.g., 'Conformism: 2' maps to a specific 'Rebel' instruction block).

Stateless JIT Compilation

The persona is not a persistent session. It is recompiled fresh for every cognitive turn. Changing a parameter mid-debate instantly alters the agent's next thought process.

Kernel vs. Persona Enforcement

User configuration (Personality) is always wrapped in an immutable System Kernel (Protocol). Users design the agent's mind, but the System enforces the 'Dirty Realism' format and cognitive discipline.

AI Team Generator

Structured Team Synthesis

Assistant accepts abstract tasks and proposes 1-N experts with reasoning. Upon confirmation, it uses internal tools to generate precise JSON configurations (archetypes, weights, traits).

Context-Aware Activation

Generator is available both pre-debate and mid-debate. The assistant accesses the active Brief to generate experts that specifically fill missing perspectives in the current discussion.

Concierge & Brief Pipeline

Concierge Pre-processing Layer

A dedicated dialog layer processes user intent first. It actively asks refining questions to disambiguate the goal before any debating agent is spawned.

Immutable Brief Anchor

Dialogue is synthesized into a structured Brief — a semantic anchor injected into every agent turn. This prevents drift and isolates the 'Goal' from the 'Conversation'.

Reliability Principles

  • Prompt InjectionAcknowledged risk. User input is embedded, but system protocols are appended last to maintain control. Trust boundary: model is trusted for content, not for format or control.
  • Token Cost ExplosionDebate depth grows costs non-linearly. Mitigated by strict sliding window truncation.
  • Stream FragilityParsing logic breaks if model outputs malformed tags. State machine includes RAW fallbacks to prevent UI crashes.

Invariants (Strict Rules)

  • ::Agent is stateless data contract (no running process)
  • ::No memory outside truncated context (MAX_MSGS=30)
  • ::Persona = Recompiled System Prompt every turn
  • ::Strict Output Protocol: [THOUGHTS]...[RESPONSE]
  • ::Client-side State is Truth (Server is Stateless)

Data Entities

users/{userId}customExperts (persona definitions)briefs (context injection)discussions/{id} (metadata)runs/{runId} (full transcript)

UX Philosophies

  • Streaming-First UI design
  • Cognitive Discipline > Speed
  • Transparent Reasoning Trace
  • No Flicker Appending

Future Signals

  • Tool calling / Execution Agents
  • Long-term Agent Memory
  • Multi-tenant API Keys
  • Autonomous Agent Loops