Dexter.cash AI Trend Analyst
Budget: $30 – $250 USD
I want to deploy a self-contained AI agent inside Dexter.cash that concentrates exclusively on financial analysis—more precisely, generating short-, mid- and long-range market trend predictions. The agent has to combine three complementary data streams so its signals are genuinely actionable:
• Historical price data
• Social media sentiment
• Economic indicators
Your job is to design, train and integrate the model, then wire it to Dexter.cash so the platform can call the agent through its existing API endpoints. A clean Python stack that may include TensorFlow, PyTorch or an equivalent framework is fine as long as the code is readable, fully documented and easy for me to redeploy.
The finished agent should:
– Pull the data sources above on a scheduled cadence, handle missing or noisy inputs gracefully and store processed features.
– Produce probability-based forecasts (directional moves, volatility ranges, confidence scores).
– Return results through a JSON response that Dexter.cash front-end widgets can consume instantly.
– Trigger webhook alerts whenever forecast confidence breaches a configurable threshold.
I’ll consider the project complete when I can run a single command to spin up the service, see fresh predictions arrive inside Dexter.cash, and view the accompanying documentation that explains model architecture, data pipelines, and retraining steps.
• Historical price data
• Social media sentiment
• Economic indicators
Your job is to design, train and integrate the model, then wire it to Dexter.cash so the platform can call the agent through its existing API endpoints. A clean Python stack that may include TensorFlow, PyTorch or an equivalent framework is fine as long as the code is readable, fully documented and easy for me to redeploy.
The finished agent should:
– Pull the data sources above on a scheduled cadence, handle missing or noisy inputs gracefully and store processed features.
– Produce probability-based forecasts (directional moves, volatility ranges, confidence scores).
– Return results through a JSON response that Dexter.cash front-end widgets can consume instantly.
– Trigger webhook alerts whenever forecast confidence breaches a configurable threshold.
I’ll consider the project complete when I can run a single command to spin up the service, see fresh predictions arrive inside Dexter.cash, and view the accompanying documentation that explains model architecture, data pipelines, and retraining steps.