Skip to content

Latest commit

 

History

History
107 lines (89 loc) · 5.2 KB

pitch.md

File metadata and controls

107 lines (89 loc) · 5.2 KB
Raw
marp
true

Multiuser Trade News: Sentiment analysis (MTNSA)

Fullstack Web3 Application.

MTNSA: serves as a helpful tool for traders and investors. It empowers users to make informed decisions by analyzing market sentiments derived from news. 📲 💹 💸

Teledev: Aleksey Savin, Alexander Badaev

BOC Fintech Hackathon 4.0, Bank of Cyprus, 2023

About us 🔎

Our team:

👨‍💻 👨‍💻 What we do: 👇
Now, the Teledev team consists of Alexandr and Aleksey. We've been working together on the 'Multiuser Trade News Sentiment Analysis' project for a while now, and our collaboration is off to a great start.

Problem 🤷

1. Market Gap 🕺: Our platform offers real-time trade news and sentiment analysis, addressing the need for timely market insights.

2. Customers 🦹‍♀️: We serve both experienced traders and novice investors with data-driven insights to enhance their strategies.

3. Costs 🧙‍♂️: We prioritize competitive pricing for high-value services.

4. Usability 👨‍💼: Our user-friendly interface enables traders at all levels to access our tools effortlessly.

5. Financials: We're committed to project sustainability through strategic revenue generation and cost management for lasting user value.

Solution 💆‍♂️

1. Closing the Gap 🏪: We offer a complete cycle from signal acquisition to trade execution, ensuring comprehensive market coverage and timely decision-making.

2. Target Audience 💅: Our primary audience consists of traders, catering to their diverse needs and strategies in the financial markets.

3. Cost Saving ☕️: We excel in analyzing large volumes of data and conducting contextual and methodological document-based queries, leading to substantial cost savings.

4. Easy to Use 📲: Our platform is user-friendly; simply connect your wallet to access 100 signals. Signal tracking is blockchain-based, ensuring transparency and accountability for a seamless trading experience.

🤖 Agents - small programs - are our helpers, here are the results:

💹 💹
Cummalative returns Cummalative returns

Ai4Finance benchmarks: FinGPT Demos

Benchmarks 1

GPT Models 🗣️

FinGPT v3.2 leads in FiQA-SA performance and efficiency, with Llama2 close but costly, while BloombergGPT lags in both, requiring model choice based on needs and budget.

Weight BloombergGPT Llama2 FinGPT v3.2
FiQA-SA 0.751 0.800 0.860
Time 53 days 21 days 2 hours
Cost $2.67 million $4.23 million $65.6
Devices 512 × A100 2048 × A100 8 × A100

Benchmarks 2:

RL Models 👾

Choose based on priorities: SharpeDQN offers safer returns, ProfitDQN yields higher maximum returns for risk-tolerant investors.

Trading System Avg. Return (%) Max Return (%) Min Return (%) Std. Dev.
ProfitDQN -0.29 49.80 -30.24 5.04
SharpeDQN 1.63 31.18 -11.29 3.76
SharpeDDQN 0.59 38.66 -7.26 3.01
ProfitD-DDQN -0.38 40.07 -27.18 4.55

Mock up 1: React Web App

📸 👨‍💻
Screens 1, 2

Mock up 2: React Web App

📲 ✅
Screens 3, 4

Roadmap ⛳️

  1. Idea: 2021.01 💡
  2. Develop alfa version: 2023.07 🥇
  3. Develop beta version: 2023.10 🏝️
  4. Release? 🎯

It's exciting to see how an idea in January 2021 (Two thousand twenty-one) will transform into a tangible product by the end of the year. The question mark at the end leaves us curious and eager for the big release in the future! 🚀🌟

Our Support Service 🫶

At our service, you'll find 24/7 online support, staffed by a team of friendly designers and consultants who are dedicated to ensuring your satisfaction. Plus, we stand by our commitment with a money-back guarantee. Your peace of mind is our top priority! 😊🌟

Key Takeaways

  1. 🇩🇪 Cloud Database in Frankfurt
  2. 📄 React SPA
  3. 💱 Monetisation and returns
  4. 🎯 Target Achievement
  5. 🕸️ Incorporates a Web3 smart contract for seamless fintech operations
  6. 🧶 Many integrations

Thank you! 🙂

Thank you so much for your attention! I'm thrilled to be of assistance. Feel free to reach out to me through the following contacts:

  1. Savin A.: You can visit my website at https://alekseysavin.com 👈
  2. Badaev A.: Check out their website at https://baadev.com 👈

Don't hesitate to get in touch if you have any questions or need further information. I look forward to connecting with you! 📸