Agentatech UG
Agenta provides an open-source LLMOps platform that enables teams to build, evaluate, and monitor reliable LLM applications through integrated prompt management and observability tools.
At a Glance
- AI Startups
- Enterprise Product Teams
- LLM Developers
AI Tools by Agentatech UG
(1)
Agenta
Open Source LLMOps Platform
Discussions
No discussions yet
Be the first to start a discussion about Agentatech UG
Latest News
CI/CD for LLM Prompts: How to Build a Prompt Deployment Pipeline
Prompt Management for Non-Engineers: Empowering Product Teams
The Definitive Guide to Prompt Management Systems
Building the Data Flywheel: Using Production Data to Improve LLMs
Products & Services
An end-to-end LLMOps platform including a playground for prompt engineering, automated evaluation pipelines, and observability tools.
A centralized hub to discover, test, and deploy various LLMs from multiple providers.
Market Position
Positioned as the open-source alternative to proprietary LLMOps platforms like LangSmith or Arize, focusing on the intersection of prompt management, evaluation, and observability with a developer-first approach.
Leadership
Founders
Dr. Mahmoud Mabrouk
CEO and Co-founder. PhD in Machine Learning/AI. Over a decade of experience in ML, previously worked on AI evaluation in academia and at startups. Founded Agenta after experiencing friction in LLM application development.
Akrem Abayed
Co-founder and Managing Director. Technical background with experience in software engineering and managing high-growth tech ventures.
Executive Team
Dr. Mahmoud Mabrouk
CEO & Co-founder
PhD in ML, decade of experience in AI evaluation.
Akrem Abayed
Co-founder & Managing Director
Technical founder and managing director of Agentatech UG.
Board of Directors
Founding Story
Founded by Dr. Mahmoud Mabrouk and Akrem Abayed during the Antler Berlin residency in 2023. Mahmoud's experience in academia and startups highlighted the difficulty of evaluating and iterating on LLM prompts, leading to the creation of Agenta as a developer-first tool to bridge the gap between prompt engineering and production.
Business Model
Revenue Model
Open-core SaaS model; provides a free open-source version (MIT license) and a paid cloud version with advanced enterprise features (SSO, RBAC, SOC2).
Pricing Tiers
2 seats, 5k traces/mo, 20 evals/mo, 30-day retention.
3 seats, 10k traces/mo, unlimited evals, 90-day retention.
10 seats, 1M traces/mo, unlimited evals, RBAC, SOC2, 365-day retention.
Unlimited seats, custom volume, audit logs, BYOC, self-hosting options.
Target Markets
- AI Startups
- Enterprise Product Teams
- LLM Developers
- Prompt Engineering Optimization
- RAG (Retrieval-Augmented Generation) Evaluation
- Production Monitoring & Regression Detection
- Non-technical Team Collaboration on AI prompts
- Various AI-driven product teams