Liquidity depth
Rewards markets with enough depth to support more credible activity.
Owned research productSolana · Risk systems · Market intelligence
Signal without
the speculation.
SolScan/X is a working Solana memecoin research product built to make fast-moving market signals easier to inspect. It combines live DexScreener data, a transparent scoring engine, rule-based risk grading, focused discovery views, and optional AI-written analysis—without pretending a score is financial advice.
6Weighted score components
8Product routes
21Verified automated tests
0–100Explainable Winner Score
The product thesis
Market data is useful only when the reasoning stays visible.
Many token tools compress dozens of inputs into a colorful badge and ask the user to trust the result. SolScan/X takes the opposite approach. Its score is deterministic, its component weights are explicit, and the interface exposes the reasons behind a result.
The product is not a prediction engine. It is a research layer: one that helps users separate liquidity, turnover, participation, momentum, social context, age, and risk flags before deciding what deserves further investigation.
01 / Winner ScoreSix weighted signals.
Six weighted signals.
One explainable model.
The composite score runs from 0 to 100, then applies a rule-based risk penalty of up to 20 points. Each component produces human-readable reasons, so a user can see what raised the score, what reduced it, and where confidence should stop.
Turnover
Measures volume relative to liquidity instead of celebrating volume in isolation.
Buy / sell pressure
Surfaces directional participation while avoiding a one-signal verdict.
Momentum
Recognizes healthy movement and rejects obviously parabolic late-entry conditions.
Social footprint
Adds public presence as context—not as proof that a token is safe.
Survivorship
Gives established markets credit for remaining active beyond the first spike.
02 / Decision interface
Built for repeated investigation—not one impressive dashboard.
The application separates discovery into focused routes for Winners, Moonshots, High Risk, Trending, New, and Saved tokens. Users can move between a visual card grid and a compact terminal table, with the preference persisted for return visits.
- Scanner, Winners, Moonshots, and High Risk views
- Trending and newly active token discovery
- Token-detail pages with score breakdowns
- Locally saved watchlist, repriced on each visit
- Responsive card and terminal-table modes
03 / Risk before hype
The model refuses to reward a vertical chart.
Momentum scoring is deliberately bounded. A token showing a parabolic short-term move—such as a 5,696% hourly spike—receives zero momentum points and a late-entry warning instead of being promoted as a winner. This is a small implementation detail with a large product implication: the system is designed to challenge excitement, not manufacture it.
Rule-based grading also evaluates red flags independently from the composite score. AI analysis is optional and secondary; when configured, Claude or GPT can explain the evidence, but neither model controls the deterministic grade.
The system behind the interface
A complete research product, not a dashboard mockup.
A React application and FastAPI service connect the interface to live DexScreener data, normalize token and pair information, calculate scores, expose risk analysis, and support optional provider-backed verdicts. AI responses are cached for 30 minutes in process, while the rest of the experience continues to work without provider keys.
The API covers health, token lists, trending and new markets, token detail, and analysis. Twenty-one documented automated tests cover the scoring engine and API behavior. Even the sparkline is labeled honestly: it is an interpolation from 5-minute, 1-hour, 6-hour, and 24-hour change data—not fabricated OHLC history.
React 18Vite 6FastAPIDexScreenerDeterministic scoringOptional Claude + GPT
04 / Product discipline
Dense information still needs hierarchy.
The mobile experience preserves the same score, risk, market, and evidence hierarchy without shrinking a desktop table into illegibility. Actions stay reachable, the card system becomes linear, and the strongest signal remains visible before supporting detail.
That design discipline matters beyond crypto. SolScan/X demonstrates how Next Design approaches any data-heavy product: define the decision, model the inputs, expose the reasoning, and give complexity a usable shape.
SolScan/X is a research tool. Nothing presented by the product or this case study is financial advice or a recommendation to buy or sell an asset.Building an intelligence product?Make the reasoning
Discuss a product build