Technical Approach
The pipeline runs as a LangGraph-style agentic loop. A planner model selects tools per query and returns a JSON policy (use_vector_search, use_metrics, use_segments, use_risks, plus vector_query). Retrieved evidence is fused into one JSON blob and passed to the response model with explicit grounding instructions, then handed to an LLM-as-a-judge critic before anything reaches the analyst.
The agentic loop: a research agent gathers grounded context from document and live-data sources, a writer drafts, and an LLM-as-a-judge critic critiques and regenerates until the draft clears the rubric, with a human-in-the-loop checkpoint before anything ships.
Planner + Tool Routing
Intent-aware retrieval selection with safe fallback to all-tools mode when planner output is invalid.
Evidence Fusion
Combines transcript snippets, metrics, segment data, and risk rows into a single context package.
LLM-as-a-Judge Critic
Scores each answer against a grounding, relevance, and completeness rubric, and regenerates any draft that fails until it clears the bar.
Hallucination Controls
Low temperature and explicit prompts to avoid guessing numbers when data is absent.
The same critic-and-regenerate philosophy behind my agentic compliance work: a judge model enforces the rubric at generation time rather than trusting a single pass, so weak or ungrounded answers are caught and rewritten before an analyst ever sees them.