Coralie
Domain opportunity intelligence platform — score, rank, and route domain portfolios with AI-backed evidence.
Role
Full-stack development of the domain intelligence platform: portfolio analysis, AI scoring engine, and evidence-backed reporting.
Architecture
VueJS, NuxtJS, NodeJS, Postgres, TailwindCSS — domain analysis pipeline with AI scoring, MCDA evaluation, and report generation.
Results
Production SaaS helping domain investors and founders make data-driven decisions on their portfolios.
Context
Coralie is a domain opportunity intelligence platform for founders, domain investors, and agencies. Most domain owners have more names than clarity — some are worth building on, some should be sold, some parked, and some dropped before they drain another year of renewal budget.
The platform turns raw domain lists into ranked business opportunities using AI analysis, signal layering, and evidence-backed scoring. The goal is not to guess but to provide actionable intelligence.
Goals and Constraints
- Analyze domain portfolios and generate quality scores
- Evaluate brandability, semantic clarity, and commercial intent
- Generate tailored business ideas for each domain's archetype
- Provide transparent, evidence-backed scoring (not black-box AI)
- Support pairwise comparison between domain candidates
- Route domains toward build, sell, park, hold, or drop decisions
Solution and Architecture
The domain analysis pipeline ingests portfolios (CSV, registrar exports, public data) and evaluates each domain across multiple signal dimensions: brandability, search demand, traffic potential, commercial value, risk, and market trends.
The AI scoring engine combines these signals using MCDA (Multi-Criteria Decision Analysis) with pairwise comparison (Elo-based ranking) to produce transparent, defensible scores.
The Opportunity Atlas presents results with clear routing recommendations, evidence confidence levels, and actionable briefs that can be shared with teams or clients.
Technical Decisions
- Signal layering over single AI scores: multiple evidence dimensions (brandability, SEO, market, risk) are evaluated independently then combined
- MCDA + Elo pairwise comparison: candidates are compared head-to-head across dimensions, producing rankings that reflect relative strength
- Evidence confidence transparency: every recommendation shows how much support exists behind it — high scores with weak evidence are flagged
- Portfolio routing framework: each domain gets a practical recommendation (build/sell/park/hold/drop) rather than forced startup ideas
- VueJS + NuxtJS frontend: responsive dashboard with interactive scorecards and shareable report generation