ARA - Agentic Research Assistant

Multi-Agent System for Financial Research Analysis

A financial research assistant purpose-built for investment analysts. Powered by autonomous agent coordination, document structure preservation, and first-class citation tracking with PDF deep links.

---

Quick Start

cp .env.template .env
# Add MISTRAL_API_KEY (required) and TAVILY_API_KEY (optional)

docker compose build --no-cache
docker compose up -d

• Frontend: http://localhost:3000
• API Docs: http://localhost:8000/docs

---

Architecture

Core Capabilities

Capability	What You Get	Business Impact
Inline Citations with PDF Deep Links	Every claim traced to source: [1] opens report.pdf#page=7	Full auditability. Analysts verify AI responses against original documents in one click.
Document Structure Preservation	Tables, figures, and headings kept intact during processing	Zero data loss. Financial tables render correctly. No broken charts or split rows.
Structured Recommendation Extraction	Raw reports parsed into queryable {asset, stance, confidence, bank, date}	Enables "Show all Overweight calls on Tech" - queries impossible with standard RAG.
Autonomous Agent Coordination	Four specialist agents orchestrated in parallel or sequence based on query complexity	Faster responses. Comparison queries run simultaneously. Dependent queries chain intelligently.
Real-Time Reasoning Transparency	Agent decision-making streamed live via SSE	Enterprise-grade auditability. See exactly which agents fired and why.

---

Specialist Agents

Four domain-expert agents, autonomously coordinated:

Agent	Domain Expertise	Execution
Knowledge Base Agent	Semantic search over research PDFs with page-level citation tracking	Vector similarity with metadata filtering
Internal Views Agent	House investment stances and recommendations	Structured SQL queries
Analyst Intelligence Agent	Analyst profiles, coverage areas, track records	Profile lookup with accuracy metrics
Web Research Agent	Live market data, news, external research	Real-time API integration

Execution Patterns:

Parallel:   "Compare Goldman's view with ours"
            → Knowledge Base Agent + Internal Views Agent (simultaneous)

Sequential: "Who covers companies in this report?"
            → Knowledge Base Agent → entity extraction → Analyst Intelligence Agent

Iterative:  "Latest on German tech"
            → Knowledge Base Agent → insufficient → Web Research Agent → aggregate

---

Document Intelligence

Structure Preservation

The ingestion pipeline preserves document semantics:

Segment Type	Processing	Result
Tables	Kept intact, never split	Financial data renders correctly
Figures	Extracted with captions	Charts preserved with context
Headings	Maintained as context anchors	Section hierarchy retained
Body Text	Chunked with heading context	Semantic coherence preserved

Chunking Strategy:

• Groups sequential segments up to 512 tokens
• Tracks page ranges per chunk for citation accuracy
• Preserves table integrity across page breaks

Recommendation Extraction

Raw text automatically parsed into structured investment stances:

{
  "asset_class": "Equities",
  "asset": "US Technology",
  "stance": "Overweight",
  "confidence": "High",
  "rationale": "AI infrastructure demand driving earnings growth",
  "source_bank": "Goldman Sachs",
  "report_date": "2024-08-15",
  "page": 7
}

Enables structured queries:

• "Show all Overweight calls on Tech from Q3"
• "Which banks disagree on European equities?"
• "Compare Goldman vs JPMorgan on fixed income"

---

Citation System

Every agent returns citations with full provenance:

{
  "citation_id": 1,
  "text": "We maintain our Overweight stance on US Technology...",
  "metadata": {
    "bank": "Goldman Sachs",
    "title": "Global Strategy Weekly",
    "report_date": "2024-08-15",
    "url": "http://localhost:8000/files/abc123.pdf#page=7"
  }
}

PDF Deep Links: Click any [1] reference to open the source page directly.

Hover Preview: See source metadata (bank, title, date) before clicking through.

Cross-Agent Aggregation: Citations from multiple agents merged with unique IDs.

---

Real-Time Streaming

Agent reasoning streamed via Server-Sent Events:

Event	Content	Display
thought	Agent reasoning process	Thoughts Panel (collapsible)
token	Response tokens	Main chat area
complete	Final answer + citations	Message with source links

The Thoughts Panel provides full visibility into agent decision-making for enterprise auditability requirements.

---

Data Storage

./data/
├── documents.db                    # Document metadata & deduplication
├── reports/
│   ├── {doc_id}.pdf                # Original PDF files
│   └── recommendations.db          # Recommendations & Analysts
├── images/{doc_id}/                # Extracted figures
└── vector_store/                   # ChromaDB persistence

Schema

Documents: PDF metadata with SHA256 deduplication

Column	Type	Description
doc_id	TEXT	UUID identifier
file_hash	TEXT	SHA256 for deduplication
bank	TEXT	Source institution
asset_class	TEXT	equity, fixed_income, multi_asset
report_date	TEXT	ISO date
chunk_count	INT	Indexed chunks

Recommendations: Structured investment stances

Column	Type	Description
id	TEXT	UUID identifier
doc_id	TEXT	Source document link
stance	TEXT	OW, UW, Neutral, Long, Short
confidence	TEXT	High, Medium, Low
rationale	TEXT	Investment thesis

Analysts: Internal analyst profiles with accuracy tracking

Vector Store: ChromaDB collection with metadata: doc_id, bank, asset_class, page_start, page_end, segment_types

---

Tech Stack

Layer	Technology	Purpose
Agent Framework	LangChain 0.3.x	ReAct pattern with autonomous coordination
LLM	Mistral Large	Query analysis, planning, synthesis
Vector Store	ChromaDB	Semantic search with metadata filtering
Structured Store	SQLite	Recommendations, analysts, documents
OCR	Mistral OCR API	Structure-preserving PDF extraction
Frontend	React + Vite + Tailwind + shadcn/ui	Streaming UI with citation hover
Deployment	Docker Compose	One-command reproducible setup

---

Design Principles

Principle	Implementation
Document Structure Preservation	Tables and figures kept intact. Financial reports have structure that matters.
Verifiable AI Responses	Page-level citations let analysts verify every claim against source documents.
Hybrid Retrieval	Vector search for semantics, SQL for structured filters. Both needed for financial queries.
ReAct Reasoning	Industry-standard agentic pattern (Yao et al., 2022) with explicit, auditable decision traces.
One-Command Deployment	Docker Compose packages everything. No manual setup steps.

---

Project Structure

├── backend/
│   ├── services/
│   │   ├── agent_orchestrator.py  # Autonomous agent coordination
│   │   ├── tools.py               # Specialist agent implementations
│   │   ├── chunker.py             # Structure-preserving chunking
│   │   ├── document_reader.py     # OCR with segment classification
│   │   ├── recommendations.py     # Structured extraction
│   │   └── database.py            # ChromaDB + SQLite
│   └── prompts/                   # Agent system prompts
├── frontend/
│   ├── src/components/
│   │   ├── Message.tsx            # Citation hover previews
│   │   ├── ThoughtsPanel.tsx      # Agent reasoning display
│   │   └── Sidebar.tsx            # Document library
│   └── src/hooks/useChat.ts       # SSE stream handling
└── docker-compose.yml

---

Future Roadmap

• Hybrid Search + Reranking (BM25 + cross-encoder)
• User Memory Layer (persistent context)
• Recommendation Backtesting (track bank accuracy)
• Multi-Modal Chart Analysis (vision models)
• Knowledge Graph (Neo4j entity relationships)
• Observability (LangSmith tracing)