Back to all case studies

Custom Agentic AI for the Tel Aviv Stock Exchange

Revolutionizing Financial Report Analysis

Industry

Finance / Capital Markets

Services

Agentic AI DevelopmentData StructuringNatural Language ProcessingCustom Software

Technology

Proprietary AI Engine, Python, OpenAI API

TL;DR

  1. 1.AI-based system for parsing quarterly financial reports
  2. 2.Natural language querying over complex financial data
  3. 3.Reduced analysis time from weeks to minutes
  4. 4.Streamlined workflow for financial analysts at TASE
TASE financial analysis interface
Tel Aviv Stock Exchange

The Challenge

TASE (Tel Aviv Stock Exchange), the cornerstone of Israel's capital markets, manages an immense volume of financial disclosures from listed companies. Each quarterly report arrives in a different format, filled with intricate financial data and nuanced terminology.

Manual review of these reports by analysts was time-consuming, inconsistent, and error-prone. Answering a single detailed financial query often required hours of cross-checking and interpretation. With the frequency and volume of reporting increasing, this process was no longer sustainable.

  • 1

    Inconsistent Report Structures

    Financial reports varied significantly in format and terminology, making automated extraction of key figures and trends virtually impossible.

  • 2

    Delayed Insight Generation

    Manual review created long delays in turning raw financials into actionable insights, which impacted strategic decision-making timelines.

  • 3

    Overwhelmed Analysts

    TASE's research teams faced an unscalable workload, limiting their ability to focus on high-value analysis and capital market development.

TASE solution overview

The Solution

AI-Powered Financial Data Extraction

We built a robust AI engine tailored to recognize and interpret financial data across various formats and structures. The system parses balance sheets, income statements, and management commentary, transforming them into a structured, queryable dataset.

Natural Language Financial Querying

Analysts can now ask complex questions—such as 'What were the main drivers of net income growth for Company X in Q2?'—and receive immediate, data-backed responses. The system supports multi-report comparisons and historical trend tracking.

Tailored Model Training

We fine-tuned the AI on TASE-specific reporting standards and local financial terminology to ensure high precision and relevance in output.

Success

TASE now operates with a significantly more agile and responsive analysis capability. What once took weeks now takes minutes. The solution streamlined the analyst workflow, reduced operational bottlenecks, and enhanced the quality of financial insights.

By automating the most tedious and complex part of the review process, TASE strengthened its position as a technologically advanced exchange and empowered its teams to focus on deeper, value-driven market analysis.

"Our collaboration with Dofinity.AI involved several months of comprehensive AI training and a targeted Proof of Concept, enabling us to unlock the powerful potential of enterprise AI. This valuable experience provided us with a clear vision of the technology's capabilities, helping us define strategic boundaries and confidently map our next steps toward innovation."

Ishai Green

Ishai Green

Head of the Data Department, Tel Aviv Stock Exchange

Need a partner for your next agentic deployment?

Let's build the agent that moves your business forward.