get it

Meeshkan allows data science teams to train, test, and deploy ML models right from Slack. Meeshkan is free and can be installed into your Slack workspace in less than a minute.

Around the web


Neil S W Murray
Alexander Pihlainen
 +14 reviews
  • Pros: 

    Esy to get started with, requires minimal adjustments to the code. Easy to integrate. Nice chatbot. Adding conditions is a great feature.


    Requires you to submit your jobs through their client.

    I have been trying out the meeshkan client in a small project of mine and I am starting to enjoy the feeling of communicating with my deep learning models that the client provides. Although I am not a big fan of not directly controlling the execution of my own scripts, the ability to submit several jobs and terminate each one manually over slack as if i was chatting with a (slightly socially awkward) person makes it well worth it.

    Mårten Nilsson has used this product for one month.


You need to become a Contributor to join the discussion.
Mike SOLOMONMaker@mikesol · ChatOps for ML at Meeshkan
Hi! Awesome to be on Product Hunt. I'm the C.E.O. of Meeshkan and it has been a wild ride getting this product out there. Like many startups, we pivoted several times before getting to this, and I would say the single most important thing that brought us here was the 50+ interviews we did with data scientists and machine learning engineers to figure out what they _really_ want to improve their workflows. Training ML models is an interactive process with a lot of teamwork & stopping/starting of models, so if teams are making these decisions on Slack, why not bring ML there too? The response has been outstanding, with our enthusiastic community suggesting great features and us working hard to implement them. If you are a data scientists or ML engineer, try it out, and if you are not on Slack, join our team! We run models all the time and would be happy to get you up and running directly on Slack. Ping for that. Also, a big THANK YOU to our partners. We are pairing up with Microsoft for Startups to do some amazing workshops around Azure ML and Meeshkan (thanks Amy Boyd, Oki Tåg & the whole MS Team!). We would also like to thank the amazing AWS team for building SageMaker and for their generous partnership with us, which allows us to expose their APIs to our community through the Slack bot ( Thanks to Bijal Nagrashna, Andjela Kusmuk, Mackenzie Kosut and Jose Manuel Rodriguez Martinez for helping us along the way! Lastly, thanks so much to Invidia inception, that helps us bring the most cutting-edge technologies into our GPU cluster management, effortlessly done right through Slack. Thanks to Serge Lemonde, Oscar Guerra and their entire team. And thanks to Stockholm AI as well for their generous feedback. We'll see you on Meeshkan - let's do some machine learning together!
Idan Tene
Idan TeneMaker@idantene
Hey Product Hunt! I'm a machine learning engineer, data scientist and software development at Meeshkan. We like to take on many roles, it keeps things interesting. Our product has recently gotten lots of attention which is amazing 😁 We're recent winners of "Slush 100", and were also recently presented on TechCrunch. Check those out if you have a moment! We also secured more partnerships with NVIDIA, AWS, Microsoft and Slack. We're booming 😁 But most importantly, we develop a product for YOU, and so we really appreciate feedback, suggestions, requests, bugs and any random thought or question (check our github repos for more tech-oriented stuff). Ping us at and I'd be personally happy to assist with whatever I can at Let's change the paradigm of machine learning. Stop pulling for results, we'll push these for you and allow you to modify your models on the fly.
Kimmo Sääskilahti
Kimmo SääskilahtiMaker@kimmo_s · Developer, machine learning researcher
Hi! I’m Kimmo, I work as ML engineer at Meeshkan. Are you still reading log files to see if your ML model is still running? Do you need to go back to your laptop many times a day to see if a training job is actually making real progress? Have you found out 24 hours too late that a big training job diverged and crashed? Meeshkan lets you develop without fear. Meeshkan monitors your jobs under the hood, letting you know the things you want and need to know, in real-time. The end result? You getting ML models to production faster and having fun along the way. Developing tools for machine learning is great, but what is even greater than developing a smart chat bot for Slack or optimizing machine learning algorithms? It’s getting feedback from our users, so please open up an issue in GitHub, ping us at or talk to me directly at! Let’s make machine learning interactive!
Lilit@lilit_antonyan · Manager
Mike SOLOMONMaker@mikesol · ChatOps for ML at Meeshkan
@lilit_antonyan Thanks for checking out Meeshkan!
Francesco Corea
Francesco Corea@francesco_ai
Hey guys, great work! I think what you are doing is paramount to the development of a new class of efficient ML models that can partition data streams across distributed infrastructure!
Mike SOLOMONMaker@mikesol · ChatOps for ML at Meeshkan
@francesco_ai thanks so much for this! Rarely do we get comments about what we are doing under the hood, but yes, a big part of our research is about this. Don't let the cute cuddly Slack bot fool you - we're putting some serious horse power under the hood!