The problem
Retail crypto investors track a lot of scattered signal with no single place that turns it into a consistent, comparable score across assets.
Social chatter, news coverage, on-chain flows and raw price action each tell part of the story - but reading all four for every asset you're watching doesn't scale. CryptoSense AI collapses them into one composite score, ranks every tracked currency by it, and keeps the reasoning visible rather than hiding it behind a single number.
- A single, comparable investment-likelihood score (0–100%) per currency
- Visible reasoning - which sources drove the score, and why
- A ranked list so opportunities are easy to scan
- A 7-day trend forecast shown as a range, not a false-precision number
Live prototype
Interactive - tap a currency to open its detail view. Data is simulated for this demo; see the disclaimer inside the app.
Scoring methodology
Each currency gets three 0–100 sub-scores, blended into one composite score and mapped to a plain-language label.
| Signal | Weight | What feeds it |
|---|---|---|
| Social sentiment | 40% | Volume-weighted NLP sentiment across X, Reddit, Telegram, with bot/spam filtering applied first |
| News sentiment | 35% | NLP sentiment and relevance scoring over licensed news APIs and RSS feeds |
| On-chain / market signal | 25% | Wallet accumulation, exchange in/outflows, volatility, liquidity depth |
A hype-vs-substance penalty is subtracted when social sentiment sharply outpaces news and on-chain signal - a proxy for speculative activity with limited underlying support - before mapping to the final label:
| Score range | Label | Interpretation |
|---|---|---|
| 75–100 | Strong buy signal | Broad, consistent positive signal across all three sources |
| 60–74 | Buy signal | Net positive signal, some mixed inputs |
| 40–59 | Hold / neutral | Signals roughly balanced or conflicting |
| 25–39 | Caution advised | Net negative signal, elevated uncertainty |
| 0–24 | Avoid signal | Broadly negative signal across sources |
System architecture
End-to-end flow from raw source data to the mobile client.
- Data sources: social platform APIs (X, Reddit, Telegram), licensed news APIs/RSS, blockchain nodes and indexers, exchange market-data feeds
- Ingestion layer: stream collectors for high-frequency sources and scheduled batch collectors for news/on-chain snapshots, buffered through a message queue into a raw data lake
- NLP sentiment pipeline: transformer-based classifiers (FinBERT-style) with entity linking, topic tagging, and bot/spam filtering
- Forecasting models: per-asset time-series models producing a 7-day expected path with confidence bands, retrained on a rolling schedule
- Scoring engine: weighted composite formula, hype-penalty adjustment, versioned score history for auditability
- API layer: REST/GraphQL service exposing rankings, per-asset detail, and historical/forecast series
- Mobile app: React Native client rendering the rankings and detail screens shown above
Tech stack
| Layer | Suggested technology |
|---|---|
| Mobile client | React Native (iOS/Android), chart library for price/forecast views |
| API layer | Node.js or Python (FastAPI), REST/GraphQL, Redis cache |
| Streaming and queue | Apache Kafka or a managed Pub/Sub service |
| NLP / ML serving | Python, HuggingFace transformers (FinBERT-style), TorchServe/BentoML |
| Forecasting | Prophet, or an LSTM/temporal-transformer model per asset |
| Storage | PostgreSQL or TimescaleDB for time-series; S3-compatible object storage for the raw data lake |
| Infra | Containerised services (Docker/Kubernetes), managed cloud, CI/CD pipeline |
Roadmap
| Phase | Scope |
|---|---|
| MVP | Top 10–20 assets, rankings + detail screens, daily-refresh sentiment, static 7-day forecast |
| V1 | Real-time sentiment refresh, push alerts on score changes, personal watchlists |
| V2 | Portfolio-aware scoring, backtesting dashboard showing historical score accuracy |
| V3 | Multi-language source coverage, community-verified source credibility ratings |