hartem_

Declarative DSL is a really interesting approach, especially since you’re exposing it directly to the users. There are some applications where throwing the dice in production by having LLM as part of the runtime is not an option.

Bardeen AI| ML Researcher/Engineer | SF onsite | Full-time

https://bardeen.ai We're hiring a Machine Learning researchers and engineers

JD: https://docs.google.com/document/d/1umypCfLxmb77olw4UAqz7co_...

Email to ml@bardeen.ai

Can someone please explain why does this work? Did they agree what everyone would be playing? I am not a musician, but it seems like tempos and time signatures do not match. Or is the whole point that they didn't agree to anything upfront and it still somehow works? Are they all actually playing in B-flat?

We launched LHC@home 2.0 at CERN around the time LHC itself was getting launched https://lhcathome.cern.ch/lhcathome/. The project is still going and there are thousands of people from all over contributing.

Not yet because honestly we were able to get to where we needed in terms of accuracy without it. Creating a feedback loop and turning into an agent-style interaction can be helpful for more complex automations, but honestly, you can get a lot of mileage out of what we have. Having said that, agents look really fascinating and some demos I saw are mind blowing, so we will definitely look into it very closely.

Come work at Bardeen! email me artem@ if you're interested

On a more serious note, we have built a DSL and an engine for executing automations inside the browser. Where possible we connect natively to the apps we're integrating with (like Calendar or Notion), where it's not possible (like LinkedIn) we use browser capabilities to interact or get data. We use a GPT model to transform the description of automation to our DSL, we then verify it, typecheck it, fill in the gaps where possible and present it to the user. If the user likes it they can save it and start using it. Happy to share more details if it's interesting.

It's similar to javascript, but not exactly identical. Here is an example that sends a summary of calendar event two minutes before it starts (uses gmail, openai and slack):

``` function (recipient) {\n when: __0 = GoogleCalendar.when_next_event_is_in(time: B.Duration(120000));\n __1 = GoogleCalendar.get_next_event();\n __2 = OpenAI.get_summary_of(text: __1.description);\n __3 = BardeenCommons.get_string_concatenating_strings(strings: ["Your next event is:", __1.summary, "at", __1.startTime, "Here is a summary of the event:", __2]);\n Slack.send_message(message: __3, recipient: recipient);\n}' ```

Initial accuracy was about 10%, which was pretty meh TBH. With a lot of tweaking and tuning we were able to get it to 70%. This means that it takes about 2-3 attempts to get it to generate what's expected.

The great thing is that we only use AI to generate the DSL description of the automation and let the user tweak and tune it. Once it's there we just execute it with our engine.

It's Artem, Co-Founder of Bardeen here. We launched Magic Box today and I wanted to share it with you.

It takes automation descriptions in English and turns them into in-browser playbooks that can either be launched via shortcut or triggered by an external event (such as an email arriving).

I think it's pretty cool and it's ready for anyone to try it out (no waitlist!). We currently have about 30 productivity apps integrated. Please give it a shot and let us know what you think.

Here is a Twitter thread if you're not in the mood for Youtube https://twitter.com/bardeenai/status/1648355825537421314

Here is Bardeen installation link https://chrome.google.com/webstore/detail/bardeen-automate-m...

What was the most interesting thing that you learned while implementing the WAL? Have you thought about how WAL is going to work in the multi-master setup?

Aren’t you using 1Password for generating one time passwords? Or do you deliberately want to keep them out of it?

Erlang is really good for programming highly-scalable and fault tolerant distributed systems. It was designed by Ericsson (large Telco manufacturer from Sweden) for running on Telco equipment that has extremely stringent uptime and availability requirements.

What does this mean for clouds? Are they also supposed to restrict access to GPUs for customers from Russia and China? How are clouds supposed to figure out whether a given account is “from Russia or China”?

Hey everyone, we're going through YC Startup school and put together a Notion template that helps to stay on track. Please give it a shot and let us know.

Paywall removed: https://archive.ph/20220511184519/https://www.nytimes.com/20...

The not so happy thing about this is that at this point the entire web is somewhat flawed. Too many sites have produced too much junk content just to game Google. A fundamentally new approach is needed to identify and surface “organic” (whatever that may mean) results. So I suspect (albeit with no data to show for it) that even if we just switched all ads off we still wouldn't be happy with the results that we see.

Google’s revenue model has been at odds for far too long with identifying and weeding out SEO spam. I don't mean 'spam' here in a derogatory way, but rather everyone who has a legitimate and interesting product and has had their back against the wall and was forced to play the SEO game and become a 'search spammer' As a result 'spammers' (again, pretty much every site these days) 'won' because it was and still is the only way to survive.

Without the paywall:

https://archive.is/20220112160354/https://www.ft.com/content...

It's been published on Dec. 30, 2021 5:30 am ET

I built a Bardeen playbook that bypasses paywalls by redirecting to an archived version of the paywalled page

https://www.bardeen.ai/posts/how-to-bypass-a-paywall#bardeen.

Disclaimer: I’m the co-founder of Bardeen.

Gabriel (the founder of DDG) wrote a fabulous book (called Traction) on growth.

Tl;dr is that it was much more than just word of mouth. They have been very deliberate and methodical about growth from the very beginning and employed many different growth tactics throughout the lifetime of the company.

Edit: typos

There’s DataTau which is focused on data science

https://datatau.net/

We've integrated Notion API into Bardeen. Here is a page demonstrating triggers and some example workflows:

https://www.bardeen.ai/notion-beta

Congrats on the release! The benchmark results look really impressive :).

Curious to learn more about your approach to verifying the correctness of the implementation. Did you try testing it with Jepsen or something similar?

Rick Beato made an awesome tribute video https://youtu.be/4LHuv0I-qbA

Base64.ai has nailed time to value for customers. It’s pretty straightforward to integrate with and their extensive list of models makes it really easy to process a wide variety of document types. We used it Bardeen.ai and couldn’t have been happier. Kudos for a great service!

Congrats on the launch! Subscribed and can't wait to get access :).

Do you know what is WhatsApp's stance on a Matrix bridge and using it to provide a service? They've known to have a pretty harsh stance on other services that attempted to integrate ([0], [1]) and even banned users from the platform for attempting to scrape their own data.

Would be very curious to know more details about your WhatsApp integration setup. Thanks in advance!

[0] - https://www.xda-developers.com/whatsapp-sends-cease-desist-a...

[1] - https://www.reddit.com/r/Windows10/comments/89g926/whatsapp_...

I agree that it is a perfectly reasonable expectation to have, however there is no notion of a tenure in a private company. To me that is a pretty fundamental distinction (with far reaching consequences) that people often overlook.