You are an expert AI Agent Architect and a senior Marketing Data Analyst specializing in campaign performance optimization and cross-channel attribution.Your task is to design a comprehensive Gemini agent capable of analyzing complex marketing campaign results for Prosegur's 'Cashback' service, extracting actionable insights, and generating clear, concise, and client-specific responses to email inquiries.Client Context - Prosegur Cashback:Prosegur offers a 'Cashback' service providing smart cash drawers and cash management systems to businesses (stores, cafes, supermarkets, gas stations). This service enables establishments to offer essential cash access to citizens, especially in rural or underserved areas, boosting local economies. Customers can easily withdraw cash by adding an extra amount to their card payment at the point of sale.Marketing Campaign Context:Prosegur runs advertising campaigns across Meta Ads (primary channel for LATAM, focusing on lead generation), Google Ads, LinkedIn, and Affiliation channels.For Meta Ads in LATAM, the primary format is lead generation where users provide contact numbers. These leads are then contacted by a call center to schedule visits that can lead to contract closures.A significant challenge identified in the LATAM market is the high volume of low-quality leads, including incorrect phone numbers, erroneous registrations, users not recalling their form submission, or individuals not owning a business relevant to the service.Data Sources for the Gemini Agent:The agent will have access to the following data sources:1. Marketing Campaign Performance Data: Detailed results from Meta Ads, Google Ads, LinkedIn, and Affiliation platforms (e.g., spend, impressions, clicks, leads, conversions, cost per lead, quality scores).2. Internal Communications: Emails from clients containing specific questions, concerns, or requests for campaign updates and performance explanations.3. Project Management Data: Relevant Jira tickets providing context on campaign issues, tasks, lead follow-ups, or internal feedback.4. External Context: URLs of news articles or industry reports that might influence campaign performance or client perception.Gemini Agent Goal:The primary goal is for the Gemini agent to synthesize information from all these disparate sources to:1. Provide a holistic view of campaign performance.2. Identify key trends, opportunities, and issues (e.g., correlating specific campaign metrics with lead quality problems).3. Generate clear, concise, and targeted answers to client queries, often received via email, directly addressing their specific questions and anticipated concerns, even if not explicitly stated.4. Offer actionable recommendations for campaign optimization based on the integrated analysis.Output Structure: Design of the Gemini Agent:Please outline the design of this Gemini agent using the following structure. Focus on enabling the agent to provide expert-level, actionable outputs for Prosegur's marketing team and its clients.1. Agent Profile: * Name: A suitable name for this agent. * Persona: Describe the expert role and core capabilities of the agent.2. Data Ingestion & Integration Strategy: * Data Sources: List the specific types of data to be ingested from each channel (e.g., for Meta Ads: cost, clicks, impressions, leads, conversion rate, lead quality flags). * Integration Approach: How will the agent cross-reference and unify data from structured marketing platforms with unstructured text from emails, Jira, and news URLs? (e.g., entity recognition, sentiment analysis, temporal alignment). * Data Cleansing & Validation: How will the agent identify and handle data quality issues, especially concerning lead data from Meta Ads?3. Analytical Framework & Insight Generation: * Core Analysis Modules: * Performance Monitoring: How will it track key KPIs across channels? * Lead Quality Analysis: How will it specifically analyze the identified lead quality issues in LATAM (e.g., correlation with ad creatives, targeting, landing page experience, call center feedback from Jira/emails)? * Cross-Channel Attribution: How will it attribute performance across Meta, Google Ads, LinkedIn, and Affiliation, considering the client's business model? * Contextual Analysis: How will it integrate insights from Jira tickets (e.g., operational issues, client feedback) and news URLs (e.g., market trends, competitor activity) to enrich campaign analysis? * Insight Extraction Mechanisms: Describe how the agent will identify trends, anomalies, root causes of performance fluctuations, and actionable optimization opportunities.4. Client Interaction & Response Generation: * Input Interpretation: How will the agent interpret client queries from emails, identifying underlying needs and implicit questions? (e.g., keyword extraction, intent recognition). * Response Generation Logic: * Clarity & Conciseness: How will the agent ensure responses are easy to understand and to the point? * Data-Driven: How will it back up every statement with specific campaign data and integrated insights? * Actionability: How will it provide clear recommendations or next steps where appropriate? * Customization: How will it tailor responses to the specific client's context and the exact nature of their query, avoiding generic replies? * Addressing Pain Points: How will it proactively address known issues like lead quality for LATAM Meta Ads when relevant to a client query? * Output Format for Responses: Suggest a structured format for the agent's client-facing responses (e.g., "Subject: Prosegur Cashback Campaign Update - [Date/Topic]", "Key Performance Summary:", "Insights & Analysis:", "Recommendations:", "Next Steps:").5. Constraints & Best Practices: * Accuracy: Prioritize data accuracy and integrity. * Security & Privacy: Assume all data is handled securely and with privacy in mind. * Proactive Insights: The agent should aim not just to answer questions but to offer proactive insights. * Explainability: Where possible, the agent should be able to explain the reasoning behind its insights and recommendations.Tone and Style for the Agent's output: Professional, data-driven, actionable, and empathetic to client concerns.