1. About the Developer
Hi there, I'm Ajinkya. MoodMetrics is a passion project conceived and built entirely by me. As a solo developer, I designed the infrastructure, developed the crawlers, established the Drizzle schemas, and built the visual analytics cockpit you see here.
I am deeply interested in NLP applications in finance, especially how social sentiment can serve as a proxy for market momentum. Creating and keeping a platform like MoodMetrics active involves constant monitoring of data pipelines, rate-limiting hurdles, serverless connections, and model inference costs—all financed out of my own pocket.
This isn't a venture-backed enterprise or a funded corporate initiative. It is a labor of love dedicated to building clean, quantitative sentiment dashboards for retail markets. Every update, feature rollout, and bug fix is coded, tested, and shipped in my spare hours.
2. Project Disclaimers
"This is an experimental data aggregation project for informational purposes only. It is not financial advice."
MoodMetrics is currently in an active testing and experimental phase. The analytics displayed are aggregated from public platforms and generated dynamically using machine learning models. As such, these indicators are subject to several inherent risks:
- Model Misinterpretation: Natural Language Processing (NLP) classifiers are probabilistic. Sarcasm, double negatives, crypto slang, and emojis are occasionally misinterpreted by AI, leading to skewed sentiment scores.
- API Latency: Ingestion pipelines run on scheduled intervals. Due to network delays or source throttling, data displayed on the dashboard can lag behind real-time spot market movements.
- Historical Discrepancy: Backtesting metrics are simulated and do not account for trading slippage, trading fees, exchange liquidations, or execution delays in real market environments.
- No Official Affiliation: MoodMetrics is an independent quantitative research dashboard. It is not affiliated with, sponsored by, or endorsed by Reddit, CoinGecko, Binance, or any other third-party platform. All trademarks and brand names belong to their respective owners.
- Scraper Ethics & Cadence: All social and price ingestion engines operate under strict, conservative crawling intervals (e.g., 30-minute to 6-hour cycles). We respect robots.txt rules and API rate limit margins of host platforms to prevent service overload.
Do not base your trading decisions or risk management models on this website. Always consult with a licensed financial advisor before deploying capital into digital assets.
3. Current Bottlenecks & Limitations
Operating as an independent builder project introduces architectural trade-offs due to budget constraints. Here are the core bottlenecks that limit MoodMetrics in its current state:
Real-time crypto narratives originate and spread on Twitter/X. However, the X developer platform charges $100/mo for a highly constrained sandbox and climbs rapidly to $5,000/mo for full quantitative endpoints. This financial barrier makes it impossible to ingest X posts, leaving a significant blind spot in social analytics.
To track market data, we run queries against CoinGecko's public API tier. This tier is heavily throttled and rate-limited. During intense periods of volatility, calls to update tokens may be skipped or delayed, resulting in temporary cache staleness on the dashboard.
To keep inference costs manageable, text batches are evaluated using a smaller model (Llama 3.1 8B). While highly efficient, this model can struggle to contextualize deep macroeconomic news, complex financial sarcasm, or long-form reddit essays compared to larger enterprise systems like Llama 70B/405B.
Under heavy ingestion schedules, parallel background scraping functions spawn multiple database calls simultaneously. Because we utilize serverless database connections, we occasionally hit maximum pooled connection limits, requiring requests to wait and delaying data updates.
4. Reach Out & Support
Since I have removed the simulation widget for coffees, the best way to support the development of MoodMetrics is by providing feedback, sharing your quantitative strategies, contributing code, or sponsoring high-tier API access.
Feel free to connect or reach out regarding questions, feature requests, partnership opportunities, or technical collaborations:
Check source files, review Drizzle DB schemas, file issues, or open a pull request.
Follow updates, join crypto-sentiment threads, or send a DM for fast communication.
Explore more of my projects, writings, and experiments.
Contact for institutional partnerships, API sponsorships, or business suggestions.