Putting The "Why" in Waii
We've shipped our most significant capability expansion yet: Deep Research. This isn't just another feature – it's the reason we called ourselves Waii in the first place.
What's New
You can now ask complex "Why" questions and get comprehensive, multi-layered answers. While our existing capabilities handle "What" questions brilliantly (What were sales last quarter? What's the trend in customer churn?), Deep Research tackles the harder problem: understanding causation, correlation, and the story behind your data.
How It Works
Deep Research operates as an intelligent agent that approaches complex questions systematically:
Question Clarification: The agent starts by understanding exactly what you're asking. "Why did sales grow in LATAM in Q1?" becomes a structured investigation with defined parameters and scope.
Multi-Path Analysis: Instead of running a single query, the agent explores multiple hypotheses. It might examine seasonal patterns, marketing campaign effectiveness, competitive landscape changes, and regional economic factors – all in parallel.
Iterative Investigation: The agent doesn't stop at surface-level answers. It follows logical threads, asks follow-up questions of the data, and builds a comprehensive picture through multiple SQL generations and analyses.
Evidence Assembly: Each finding gets validated through additional queries and cross-references. The agent builds a case using your actual data, not assumptions.
What You Get
Comprehensive Reports: Full investigation summaries that explain not just what happened, but why it happened. Each report includes the reasoning process, key findings, and supporting evidence.
Rich Visualizations: Charts and tables that illustrate the story. These aren't just pretty pictures – they're specifically chosen to support the narrative the agent has discovered.
Complete Transparency: Access to the full event stream showing every decision the agent made, every hypothesis it tested, and every query it ran. You can see exactly how it reached its conclusions.
Technical Implementation
Deep Research is available through multiple integration points:
- API Access: Direct integration into your applications with full programmatic control
- Model Context Protocol (MCP): Seamless integration with MCP-compatible tools and workflows
- Agent-to-Agent (A2A): For complex multi-agent scenarios where Deep Research collaborates with other AI systems
The agent leverages our existing metadata understanding, SQL generation capabilities, and visualization engine – but orchestrates them in an entirely new way to handle complex, multi-faceted questions.
Real-World Impact
This capability transforms how product and engineering teams approach data investigation. Instead of spending hours or days manually exploring different angles, you can ask "Why are mobile users churning faster than desktop users?" and get a thorough investigation that might uncover UX issues, performance problems, or feature gaps you hadn't considered.
The agent doesn't just give you correlation – it actively seeks causation by testing multiple hypotheses against your data. When it can't find definitive answers, it clearly states what additional data would be needed to complete the investigation.
Integration Benefits
Deep Research maintains our core principle: it fits into your existing environment without replacement. Your current dashboards, workflows, and tools continue working exactly as before. Deep Research adds a new layer of analytical capability that your team can access when they need to go beyond standard reporting.
The agent respects your existing security models, data governance policies, and access controls. It works within the constraints of your current setup while delivering enterprise-grade analytical capabilities.
This release represents a fundamental expansion of what's possible with AI-driven data analysis. You're not just getting faster answers – you're getting deeper understanding.