News Stocks Fractal Analytics Launches PiEvolve AI Engine, Achieves Top Performance on OpenAI's MLE-BenchFractal Analytics Launches PiEvolve AI Engine, Achieves Top Performance on OpenAI's MLE-Bench
Fractal Analytics launched PiEvolve, an evolutionary agentic engine for autonomous machine learning that ranks second on OpenAI's MLE-Bench with 61.33% performance. The system achieved 47 medals (28 gold, 12 silver, 7 bronze) across 75 competitions and became the first agent to surpass 60% Overall Medal Rate and 80% MLE-Bench-Lite performance. PiEvolve features continuous optimization, intelligent memory systems, and production-ready capabilities for enterprise ML applications.
*this image is generated using AI for illustrative purposes only.
Fractal Analytics Limited has announced the launch of PiEvolve, an evolutionary agentic engine designed for autonomous machine learning and scientific discovery. The Mumbai and New York-based global enterprise AI company revealed that PiEvolve has achieved significant performance milestones on industry-standard benchmarks, positioning it among the top-performing AI systems globally.
Benchmark Performance Achievement
PiEvolve has demonstrated exceptional performance on OpenAI's MLE-Bench, a benchmark that evaluates how effectively AI systems can solve real-world machine learning challenges. The system achieved notable milestones by becoming the first evaluated agent to surpass 60% in Overall Medal Rate and 80% in MLE-Bench-Lite performance.
| Performance Metric: | Achievement |
|---|---|
| Overall Medal Rate: | 61.33% |
| MLE-Bench-Lite Performance: | Above 80% |
| Total Medals Won: | 47 |
| Gold Medals: | 28 |
| Silver Medals: | 12 |
| Bronze Medals: | 7 |
| Total Competitions: | 75 |
On the OpenAI MLE-Bench leaderboard, PiEvolve secured the second position with a 61.33% score, closely following LoongFlow at 62.66%. The system outperformed other notable agents including Famou-Agent 2.0 (59.56%), ML-Master 2.0 (56.44%), and several other competing systems.
Technical Architecture and Features
PiEvolve distinguishes itself from traditional machine learning models through its continuous improvement approach. Unlike conventional models that are trained once and deployed, PiEvolve continuously tests and improves its own solutions until the available compute budget is fully utilized. The system is built on a graph-structured search architecture that integrates reasoning, code generation, and validation within a unified iterative process.
Key technical features include:
- Continuous Optimization: Iteratively evolves candidate solutions, improving performance until computational limits are reached
- Intelligent Memory: Uses priority-based sampling with decay to avoid local optima and ensure diverse exploration of solution paths
- Dual Strategy: Improves high-performing solutions while actively debugging weaker ones to elevate overall system performance
- Production-Ready: Includes Pause and Resume capabilities for long-running workloads and integrates seamlessly into enterprise ML pipelines
- Graph-Structured Search: Systematically explores reasoning, code, and validation loops to generate and refine solutions
Efficiency and Computational Performance
PiEvolve demonstrates strong efficiency characteristics, combining high performance with optimized resource utilization. Within a standard 24-hour run, it delivers results comparable to systems requiring longer runtimes and greater compute resources. Notably, even the 12-hour version of PiEvolve (using approximately 50% of the compute budget) achieved a 52.00% score, ranking among the top-performing agents and identifying high-quality solutions early in the process.
Leadership Commentary
Srikanth Velamakanni, Co-founder, Group Chief Executive and Vice Chairman of Fractal, emphasized the significance of the achievement: "MLE-Bench is widely regarded as the gold standard for evaluating AI agents on real-world machine learning tasks. PiEvolve's ranking among the top systems globally is a meaningful validation of our research direction. At Fractal, our ambition has always been to power every human decision in the enterprise. PiEvolve advances that mission by enabling AI systems that continuously improve and deliver measurable business outcomes."
Suraj Amonkar, Chief AI Research and Platforms Officer at Fractal, highlighted the competitive landscape: "PiEvolve ranks among the top agents evaluated on OpenAI's MLE-Bench, including systems developed by leading global research labs. Achieving top-tier performance on OpenAI's MLE-Bench is a significant milestone for our research team and reinforces Fractal's commitment to advancing enterprise-grade AI."
Enterprise Applications and Market Position
The system is designed to tackle complex, multi-variable optimization problems across various sectors including supply chains, financial services, and data center operations, where static AI systems often struggle to perform at scale. PiEvolve brings continuous improvement, efficiency, and production reliability together in a single system, helping organizations solve complex challenges with confidence and scale.
With this launch, Fractal reinforces its position at the forefront of enterprise AI innovation. The company operates across global locations including the United States, Canada, the UK, the Netherlands, Ukraine, India, Singapore, South Africa, the UAE, and Australia, employing over 5,000 professionals worldwide.
17
