An Indian startup just did something nobody expected — outranked the dedicated AI research divisions of global tech giants on the most rigorous enterprise AI benchmark in the world. Bengaluru-based DecisionX has secured the #2 spot globally on Spider 2.0 Lite, and the tech world is paying attention.
Table of Contents
DecisionX Spider 2.0 Lite Performance
| Detail | Info |
|---|---|
| Company | DecisionX, Bengaluru |
| Benchmark | Spider 2.0 Lite |
| Global Rank | #2 |
| Score | 71.84 |
| Result Posted | April 29, 2026 |
| Indian Rank | #1 (Only Indian company in Top 10) |
| Beaten Competitors | JetBrains, Tencent, Samsung SDS, Snowflake, Tsinghua University |
What Is Spider 2.0 Lite — And Why Does It Matter?
Not all AI benchmarks are created equal. Spider 2.0 Lite, presented at ICLR 2025, is built from 547 real-world enterprise database problems — the kind that break most AI tools the moment they leave a controlled demo environment.
The benchmark tests an AI system’s ability to reason over database schemas with 1,000+ columns, generate multi-step SQL queries across BigQuery, Snowflake, and SQLite, and resolve genuinely ambiguous business questions with real execution accuracy. In short, it’s designed to expose systems that can’t handle the messy complexity of actual enterprise data.
For companies evaluating artificial intelligence tools for strategy and analytics, this is the benchmark that separates genuine capability from polished demos.
How DecisionX Built a #2-Ranked AI Engine
DecisionX’s edge comes from two core engineering achievements that most enterprise AI tools simply don’t have.
First — Semantic understanding before reasoning begins. Most AI tools reason cold, with no grasp of what a company’s data actually means. DecisionX builds context through three layers:
- Data Ontology — data definitions and causal relationships
- Domain Ontology — industry and category know-how
- Decision Ontology — a structured record of past decisions
Together, these form a living causal graph that grows sharper with every interaction — institutional memory that compounds over time.
Second — Specialised reasoning agents purpose-built for root cause analysis, forecasting, and optimisation, all operating through the same shared context.
As Ranjan Kumar, Founder and CEO of DecisionX, put it: “We are built for the hardest version of enterprise reasoning from Day 1. This proves it.”
The platform is already live in production across banking, healthcare, life sciences, consumer, and industrial sectors — not a prototype, a working system at enterprise scale.
Why This Is a Big Deal for Indian AI
This isn’t just a benchmark win — it’s a statement. Every other company in the Spider 2.0 Lite top 10 is either a large multinational tech corporation or a specialized academic lab. DecisionX is the only Indian company in that list, and it finished above all of them except one.
It’s a landmark moment for India’s enterprise AI ecosystem and proof that deep-tech startups from Bengaluru can compete at the very highest global level.
Explore more on India’s AI and tech rise with our coverage of top Indian AI startups making global waves on TechnoSports, our breakdown of India’s best deep-tech companies to watch in 2026, and the latest enterprise AI trends shaping business strategy.
FAQs
Q: What is Spider 2.0 Lite and why is DecisionX’s #2 ranking significant?
A: Spider 2.0 Lite is the world’s hardest enterprise database reasoning benchmark; DecisionX’s #2 finish makes it the highest-ranked Indian AI company ever on a benchmark of this complexity.
Q: Which global companies did DecisionX beat on the Spider 2.0 Lite leaderboard?
A: DecisionX ranked above the AI research divisions of JetBrains, Tencent, Samsung SDS Research America, Snowflake, and Tsinghua University.
