Inspiration Small startups struggle with hiring more than anything else. As a solo developer and early-stage builder, I’ve personally experienced how painful and time-consuming the hiring process can be. Screening resumes, scheduling calls, conducting repeated first-round interviews — it can easily take 40+ hours per hire, which slows down product development and kills momentum. Large enterprises have automation, full HR teams, and expensive tools. Small teams have… Google Sheets. HireVision AI was inspired by one simple question: “What if a 2-person startup could hire with the same power and speed as Google?” That question became the foundation of this project. I wanted to create a fully autonomous hiring assistant — one that handles resume screening, interview conversations, analytics, and intelligent decision-making, all without needing a human recruiter.
What it does HireVision AI is a complete AI hiring system built fully solo in 15 days for the AI Championship 2025. The system automates the entire first half of the hiring pipeline: Core Capabilities Autonomous AI voice interviews Real-time proctoring (face detection, sentiment, attention tracking) Smart resume evaluation and scoring Automated candidate ranking Real-time dashboard analytics Secure authentication & role-based access The platform feels simple for the user, but underneath it runs a powerful combination of Raindrop + Vultr + ElevenLabs + Gemini, connected through a production-grade Next.js application.
How we built it How It Uses Raindrop I built this project end-to-end: Designed the UI/UX Integrated Raindrop MCP Connected Vultr database + storage Implemented real-time face detection Configured Firebase Auth + WorkOS Integrated ElevenLabs streaming voice Built API routes, error handling, and pipelines Wrote matching logic and analytics Recorded, edited, and scripted the demo This was one of the most intense but rewarding builds of my life — learning nonstop, debugging late nights, and stitching systems together until everything clicked. Raindrop is the intelligence layer of the entire system: SmartSQL → AI-driven PostgreSQL queries SmartMemory → Stores conversational context across interviews SmartInference → Processes resumes and calculates match scores SmartBuckets → Secure file storage & access Raindrop allowed me to build advanced reasoning workflows without writing huge amounts of backend code. It is the “brain” of HireVision AI.
How It Uses Vultr Vultr powers the backend infrastructure: Vultr Managed PostgreSQL stores all job postings, candidates, evaluations, and analytics Vultr Object Storage handles uploaded resumes and associated documents Fast global performance ensures the app runs smoothly during interviews and scoring Vultr gave the project scalability and production reliability with minimal setup.
ElevenLabs + Gemini ElevenLabs provides the real-time, conversational AI voice that conducts interviews. Gemini 2.0 Flash powers the reasoning, follow-up questions, sentiment understanding, and structured outputs. Together, they create a surprisingly natural interview experience.
Challenges we ran into Multi-service integration Raindrop, Vultr, ElevenLabs, Gemini, Firebase — combining all these while keeping the UI fast required careful design. Real-time voice interview logic Streaming audio, generating follow-ups, and syncing analytics was complex. Smart resume scoring Designing a system that feels intelligent but also fast required experimentation. Building everything solo From backend to frontend to AI pipelines to video demo — doing it all alone in 15 days pushed me to my limits. ## Accomplishments that we're proud of Despite the challenges, HireVision AI reached a level of polish and functionality I’m truly proud of — especially as a solo build completed in just 15 days. Fully functional, end-to-end hiring automation From resume upload to AI interview to analytics — everything works seamlessly in real time. Real-time AI interviews with voice, sentiment, and proctoring The interviewer feels natural, responsive, and intelligent, thanks to ElevenLabs + Gemini + Raindrop integration. Smart resume analysis with 85%+ reliable scoring The system generates structured insights, skill matches, and ranking with impressive consistency. Production-ready backend with Vultr Managed PostgreSQL + Object Storage + secure pipelines mean the platform can scale beyond demo usage. Clean, intuitive UI built from scratch Even under time pressure, the interface remained simple, elegant, and easy for users to navigate. A fully solo-built project Every line of code, every design decision, every integration, and the entire demo was made by one person — which makes the final result even more meaningful. A unified experience that feels enterprise-grade The smooth flow between AI systems, backend infrastructure, and UI elements makes the app feel like a real production product, not a hackathon prototype. What we learned How to design and scale multi-agent systems How to integrate Raindrop Smart components effectively Real-time audio + sentiment processing Better system design, caching, and reliability patterns How to manage large projects solo under pressure Most importantly, I learned how powerful small teams (or even one person) can be when backed by great tools.
What's next for HireVision HireVision AI began as a 15-day solo project, but it has already shown the potential to grow into a full-scale hiring platform for startups around the world. Here’s what comes next:
Multi-Agent Hiring Workflow Introduce specialized AI agents that handle: Job description refinement Candidate outreach automation AI-driven shortlisting Interview summary generation Offer letter automation A complete, autonomous hiring pipeline.
Team Collaboration Features Allow multiple team members to: Review candidates together Leave comments and evaluations Share AI-generated insights Approve hiring decisions Bringing startup workflows closer to enterprise HR systems.
Advanced Proctoring & Behavioral Insights Enhance the interview system with: Emotion timeline analysis Gaze consistency tracking Attention drift detection AI-scored soft-skill evaluation Helping teams identify the strongest candidates more accurately.
Employer & Candidate Mobile Apps A lightweight mobile version of HireVision for: On-the-go interviews Instant application status updates Push notifications for hiring teams Real-time dashboards
Auto-Sourcing & AI Outreach Enable the system to: Find candidates across the web Evaluate profiles automatically Send personalized AI-written outreach messages Reducing sourcing time from weeks to hours.
Marketplace for Voice Interview Models Create an “Interview Voice Store” where: Employers choose interviewer personality Domain-specific interview models exist Candidates experience more tailored interviews
Enterprise Features Add support for: SSO integrations Audit logs Applicant tracking pipelines Compliance & data privacy controls Allowing HireVision to scale beyond startups.
Monetization & Startup Plans Introduce: Pay-per-interview Subscription plan Usage-based Raindrop credits Team-based pricing Designed to keep the product affordable for small teams.
AI-Curated Candidate Profiles Automatically generate: Skill maps Experience summaries Culture fit analysis Project breakdowns Strength & improvement areas A complete candidate dossier in seconds.
Scaling globally with Vultr With Vultr’s global data centers: Deploy regionally Reduce latency Expand to international users Provide a smoother real-time experience
Log in or sign up for Devpost to join the conversation.