Aster

Competitive Intelligence & Product Roadmap

Aster automates executable AI research experiments.

Company Overview

Aster is an autonomous research lab platform that turns a benchmark, program, and evaluator into iterative AI-led experiments. It serves AI researchers, ML engineers, and scientific teams with executable research tasks.

Latest Intel

Zeitgeist tracks private signals to determine where the company is heading strategically.

May 18, 2026

Technical Approach Signal

Confidence:

Medium

New Intel: Aster is building around benchmark-driven program evolution rather than chat. That puts it closer to OpenEvolve and AlphaEvolve-style research automation, with evaluator setup and compute cost as the key adoption gates.

What They're Building

The company's public product roadmap & what they're committed to building.

Autonomous Research Runs

Aster takes a task, starting program, and evaluator, then runs iterative experiments toward a measured result.

Hypothesis Proposal

The product surface centers on generating research hypotheses rather than serving as a chat interface for researchers.

Hypothesis Testing

Aster tests proposed changes through executable evaluators, making benchmark feedback the control loop.

AI Research First

The company frames the broader goal as AI-native research for many domains, starting with research on AI itself.

Founder and Key Execs

Emmett Bicker

Founder and CEO (former post-training researcher at Magic working on long-context coding LLMs)

Founder Force Multiplier

Emmett Bicker's Magic work on long-context coding LLMs maps directly to Aster's research-agent loop. A solo founder also means limited public evidence of commercial coverage.

Funding History

2026 | Founded
2026 | Y Combinator Spring 2026, with public funding snippet reporting $500K from Y Combinator

Competitors

OpenEvolve:

OpenEvolve is the closest benchmarked alternative, with Aster claiming faster convergence on circle packing.

AlphaEvolve:

AlphaEvolve is the larger research-lab reference point for AI systems that search over programs and algorithms.

TTT-Discover:

TTT-Discover is a research comparator that Aster reports surpassing on several benchmark tasks.

Hiverge:

Hiverge targets automated algorithm discovery, while Aster's public surface is broader executable research automation.

Intology:

Intology also frames AI systems as research workers, with Aster more visibly tied to benchmark-driven program improvement.

Aster

's Moat:

Technical infrastructure is the candidate moat: evaluated programs, metrics, and prompts could compound into better search, but public proof of shared proprietary data is thin.

How They're Leveraging AI

Benchmark-driven program evolution

Aster uses LLM agents to modify code, run evaluators, and iterate toward better benchmark scores for AI and scientific research tasks. The user supplies the research task, starting program, data or benchmark, and evaluator.

Automated research hypothesis generation

Aster generates research hypotheses for AI and scientific problems, then tests them against executable metrics instead of stopping at written suggestions. This turns hypothesis quality into something measurable.

Cross-run research memory

Aster appears to persist evaluated programs, metrics, and prompts from research runs, creating a memory layer for future experiment search. The data advantage is plausible, but shared learning across customers is not publicly proven.

AI Use Overview:

Aster applies LLMs as program-evolution agents, proposing code changes, running evaluators, and using metrics to steer the next hypothesis.

AI Developer Tools

Entire

Git-native AI code explainability and session context capture

The ex-GitHub CEO is building the compliance layer for AI-generated code, with personal relationships to every enterprise buyer who will need it.

Approxima

Lets product teams go from idea to deployed software in under an hour with AI agents.

Most AI coding tools target greenfield features. Approxima goes after the unglamorous maintenance work (bug fixes, incremental updates) that eats 60%+ of engineering time, with sandbox validation that lets agents merge to production without human review.

21st Labs

Helps developers ship AI apps 10x faster with purpose-built components and agent tools.

Alt-X

Replaces 12-hour manual modeling sessions with one prompt that builds deal models from raw docs.

Real estate underwriting still runs on 12-hour Excel sessions built from 200-page PDFs. Alt-X collapses that into a single prompt, and PE firms managing hundreds of millions in AUM are already using it.

Back To All Companies >