Homeβ€ΊCompaniesβ€ΊMagicflow 𐂂

Magicflow 𐂂

The Gen AI evaluation and curation platform

Prompt engineers, visual artists and AI data scientists are using Magicflow to run experiments at scale and perfect their pipelines Our product is ready to use: https://console.magicflow.ai try it out! we have some preloaded examples you can try. 🀟 Talk to us about anything! We love meeting new people https://calendly.com/magicflowai/
Magicflow 𐂂
Founded:2022
Team Size:2
Location:Tel Aviv-Yafo, Israel
Group Partner:Diana Hu

Active Founders

Yarden Shem Tov, Founder

Yarden likes to build stuff from scratch, he is the founder of a tech-oriented short-term rental company in Tel Aviv and of what was one of the most popular cars websites in Israel (200K+ visitors per month) I have been coding since I was 9 years old and have served as a founding team engineer for several start-ups building the MVPs
Yarden Shem Tov
Yarden Shem Tov
Magicflow 𐂂

Yotam Hamiel, Founder

Software engineer with 15+ years of experience building products, ETLs & ML infra. Ex-Flexport, Ex-Facebook, Ex-8200

Company Launches

Hey! We are Yotam Hamiel and Yarden Shem Tov, the founders of Magicflow!

Magicflow is a no-code drag-and-drop tool to design and run AI workflows that easily integrate into your app.

Complex workflows tend to be expensive and slow. We can optimize them for cost and speed so you don’t have to deal with the painful infrastructure work.

The Problem:

AI capabilities such as ChatGPT and Stable Diffusion are more accessible than ever. However, they still require a fair amount of technical know-how to deploy and orchestrate, particularly when combining multiple models to create optimized and cost-efficient workflows.

Our Solution:

Magicflow β€” a no-code AI model orchestration solution.

We provide a user-friendly playground that integrates models from top services like HuggingFace, OpenAI, and Replicate.

With Magicflow anyone can easily stitch together various models without requiring any DevOps or code.

Once ready to deploy, workflows are cost/speed optimized and made ready for production.

An example of a workflow built with Magicflow.

We came up with Magicflow by talking to our pre-pivot customers (who were mostly using us as a proxy for AI services). They told us that they needed a way to quickly iterate on their workflows, debug them and deploy them to production in a scalable and cost-efficient way.

Our goal is to provide an accessible solution, for both engineers and non-engineers, to utilize the vast AI capabilities available to enhance their own products.

Suhail (Mixpanel founder) described it best:

How you can help:

We’re looking to work with companies who rely on AI workflows, and are looking to make them cheaper and faster. If you know someone please contact us at founders@magicflow.ai