A deep dive into building a "Market Sentiment Monitoring and Alert Agent" and key takeaways from the 5-Day Google & Kaggle AI Agents Intensive.
Introduction
Before taking the 5-Day AI Agents Intensive Course with Google and Kaggle, I viewed Large Language Models (LLMs) primarily as knowledge engines—advanced chatbots that could write code or answer questions. By the end of the week, my perspective had completely shifted. I realized that when you give an LLM "tools" and a "reasoning loop," it stops being just a text generator and becomes an Agent: a system that can perceive, reason, and act.
For my capstone project, I applied these learnings to build a Market Sentiment Monitoring and Alert Agent, a system designed to automate the tedious task of tracking financial news and gauging market health.
Here is my reflection on the journey, the core concepts that clicked, and how I built my agent.
The "Aha!" Moment: Tools & Reasoning
The most resonant concept from the course was the transition from Zero-Shot Prompting to ReAct (Reasoning + Acting) loops.
Before: I would ask an LLM, "What is the sentiment of the market?" It would hallucinate or give a generic answer based on old training data.
After: I learned to build an agent that thinks:
Thought: "I need to check the latest news for specific tickers."
Action: Calls a Search Tool or News API.
Observation: Reads the headlines.
Reasoning: "These headlines look bearish. I should calculate a sentiment score."
Final Response: Sends an alert.
Giving the AI "hands" (Tools) and a "brain" (Gemini 2.0 Flash) allowed me to build something dynamic.
My Capstone: Market Sentiment Monitoring and Alert Agent
For the final project, I wanted to solve a real-world problem: Information Overload in Finance. Traders can't read every headline, but an Agent can.
What it Does
My agent autonomously monitors financial news sources, analyzes the sentiment of headlines using Google's Gemini, and triggers an alert if the market sentiment swings too far into "Fear" (Bearish) or "Greed" (Bullish) territory.
Under the Hood
The Brain: Google Gemini (via the Vertex AI or Gemini API).
The Tools: I integrated Python libraries to fetch live data (e.g., yfinance or web scraping tools) and search capabilities.
-
The Logic: The agent doesn't just read; it assigns a numerical score to the qualitative data.
- Bullish Keyword detection
- Context awareness (e.g., distinguishing between "loss" in profits vs. "loss" of data)
You can check out the full code and write-up here:Link to my Project
Challenges & Key Takeaways
Building this wasn't without its hurdles.
Hallucinations vs. Reality: I learned the importance of Grounding. Using Google Search as a tool ensured the agent wasn't making up news but referencing real URLs.
Context Windows: Managing how much history the agent remembers (Memory) is crucial. You don't want it to analyze news from 2020 when deciding on today's alert.
Evaluation: As emphasized on Day 4 of the course, "vibes" aren't enough. I had to manually check the agent's sentiment scores against my own intuition to tune the prompts.
Conclusion
The Google & Kaggle Intensive was a sprint, but it laid the foundation for a marathon. Moving forward, I plan to upgrade my Market Sentiment Agent to not just alert but potentially suggest portfolio adjustments (in a paper-trading environment, of course!).
If you are on the fence about diving into Agentic AI, start now. The gap between "talking to AI" and "having AI do work for you" is closing fast.
Top comments (0)