Inspiration
Humans tend to act more emotionally than rationally. This can lead to impulsive trading decisions based on gut feelings rather than careful analysis. A trend identification application uses data-driven insights to spot potential opportunities and mitigate risks, helping investors make more calculated and profitable choices within the cryptocurrency market.
What it does
Crypto-Trends empowers cryptocurrency investors by:
- Reducing the impact of emotional decision-making through data-driven analysis.
- Saving time by automating the manual trend identification process.
- Increasing potential profitability by helping users make informed trades based on market trends.
How we built it
The application was developed using python, golang, javascript, next.js, mongodb, rabbitmq and google-ai gemini. We are using prompt engineering to ask gemini for a technical analysis into the data send in the prompt to identify the trend.
Challenges we ran into
- Training time-series data on Gemini was a bit hard, probably due some gap of knowledge in data science field.
- Get the accuracy of trained data in Gemini.
Accomplishments that we're proud of
We aim to empower traders to make more informed, data-driven decisions, ultimately increasing their chances of success in the crypto market.
What we learned
We gained valuable experience integrating with various cryptocurrency exchanges' APIs and handling potential inconsistencies in their data formats. Also how to handle hallucinations in a model.
What's next for Crypto Trends
- medium and long term trends
- watchlist
- volume and price changes
- analysis based on different technical indicators
- wallet with performance
Built With
- coinmarketcap-api
- gemini
- golang
- google-ai
- google-cloud
- javascript
- mongodb
- next.js
- python
- rabbitmq
Log in or sign up for Devpost to join the conversation.