Unirivo
Where great minds converge
The problem Unirivo solves
Unirivo: Solving the Friction in Collaborative Innovation
Unirivo is a comprehensive project incubation hub designed to address the systemic barriers that prevent ideas from becoming reality. By integrating AI-driven management with Blockchain security, Unirivo solves the following critical problems in the modern creative and technical economy.
1. The Trust and Security Gap (IP Protection)
The Problem: Many creators are hesitant to share their innovative ideas for fear of intellectual property theft. Traditional collaboration platforms offer no legal protection at the point of entry, leaving founders vulnerable during the initial "pitching" phase.
The Unirivo Solution: Unirivo implements a mandatory, automated "No-Reveal" security gate. Before a user can view sensitive project details or join the team, they must sign a legally binding NDA within the platform. This creates a secure environment where founders can disclose details to potential talent without anxiety.
2. Inefficient Talent Acquisition and Vetting
The Problem: Finding the right collaborators is a time-consuming manual process. Reviewing resumes often fails to reveal a candidate's actual communication style or technical depth, and manual interviewing is a significant bottleneck for solo founders.
The Unirivo Solution: Unirivo uses AI to automatically scout talent based on the specific skill requirements of a project. It further streamlines vetting via an automated AI-voice interview system. The AI evaluates candidates in real-time based on role-specific criteria, ensuring only qualified and compatible members join the project.
3. The Management and Execution Trap
The Problem: Most independent or "passion" projects fail due to a lack of structure. Without a dedicated project manager, teams struggle with vague milestones, disorganized task distribution, and technical blockers that bring progress to a halt.
The Unirivo Solution: The platform acts as an AI Project Manager. It automatically generates actionable milestones and manages the task board to keep the team focused. Additionally, the community feature—backed by AI troubleshooting—ensures that technical issues are resolved instantly, preventing the project from stalling.
4. Unverifiable Proof of Contribution
The Problem: Traditional resumes rely on self-reported experience, which is often difficult for employers to verify. Conversely, contributors to small or private projects often have no official record of their hard work once the project ends.
The Unirivo Solution: Unirivo leverages the Solana Blockchain to issue encrypted Contribution Certificates. These are immutable, tamper-proof records of the specific tasks and milestones a user completed. This transforms "work experience" into a verifiable digital asset, providing workers with high-integrity proof of their skills.
5. Community Isolation
The Problem: Builders often work in silos, losing out on the collective intelligence of other creators. When teams hit a wall, they have nowhere to turn for specialized advice without hiring expensive consultants.
The Unirivo Solution: Unirivo fosters a "builders-helping-builders" ecosystem. Through the community feature, users can raise issues that the AI can answer immediately based on the project's context, or the wider community can provide peer support, creating a collaborative knowledge base for all users.
Challenges we ran into
Technical Challenges and Solutions
1. Context Management in AI-Driven Project Planning
The Challenge
As projects progressed, the volume of chat logs and task metadata grew beyond the standard context window limits of Google Gemini and Groq. Attempting to feed the entire project history into the AI for milestone generation resulted in high latency, increased token costs, and a loss of focus on the most recent project requirements.
The Solution
I implemented a tiered retrieval system using the Mastra AI Framework. The system performs periodic summarization of project activities, which are stored as "Project States" in MongoDB. When a user requests new milestones or task updates, the system only injects the current project state and the most recent relevant interactions. This ensures the AI remains highly accurate and responsive without being overwhelmed by historical data.
2. Latency in AI-Voice Interview Feedback Loops
The Challenge
A critical feature of Unirivo is the AI-driven voice interview. Initially, there was a significant delay between the user's spoken response and the AI's next technical question. This "response gap" made the interaction feel disjointed and hindered the assessment of candidate communication skills.
The Solution
To address this, I moved to a streaming architecture. Using Groq's LLaMA 3.3 for rapid text generation and a low-latency Speech-to-Text (STT) / Text-to-Speech (TTS) pipeline, I enabled the AI to begin streaming audio chunks before the full response was finished. Additionally, I utilized GSAP to animate "active listening" states on the frontend, which effectively masked the remaining processing time and provided a seamless user experience.
3. Data-Level Enforcement of the No-Reveal Security Gate
The Challenge
Ensuring the security of intellectual property was paramount. Early testing revealed that simply hiding the project dashboard behind a UI modal was insufficient; savvy users could still access sensitive project data through direct API requests to the /api/project routes before signing the mandatory "No-Reveal" contract.
The Solution
I implemented a server-side security gate using Next.js Middleware and Mongoose pre-find hooks. Every request targeting project-specific data now undergoes a mandatory permission check against the user's database record. If the nda_signed flag is not set to true, the API rejects the request at the server level with a 403 Forbidden status. This ensures that the idea is protected at the source, regardless of the frontend state.
4. Solana Transaction Finality and Certificate Issuance
The Challenge
When minting contribution certificates on the Solana blockchain, the frontend would often report a successful transaction before the data was fully propagated across the network. This caused an error when users attempted to view their certificates on a block explorer immediately after project completion.
The Solution
I refined the Solana Web3.js integration to use the finalized commitment level for critical transactions. I developed a recursive polling mechanism in the backend that waits for a cluster-wide confirmation signature before updating the certificate status in the user's dashboard. This ensures that whenever a user is notified of a successful contribution record, the link to the blockchain is 100% verified and accessible.
5. UI Synchronization with Complex State Transitions
The Challenge
With multiple moving parts—AI interviews, blockchain minting, and real-time task management—keeping the UI in sync without constant page refreshes was difficult. Standard React state management led to "prop drilling" and inconsistent UI updates across the chat and task modules.
The Solution
I utilized a combination of React 18's concurrent features and custom hooks to manage local state, while Radix UI primitives provided a consistent accessibility layer. Framer Motion was used to handle the transition between the "Recruitment" phase and the "Active Project" phase, ensuring that the heavy lifting of state transitions remained invisible to the end user.