How to Write a Strong Conference Proposal?

Recording: How to Write a Strong Conference Proposal?

thank you for joining us today for this session uh about how to write a strong conference proposal and uh I am o did I am um director of deel at BX I am also a proud Co organizer of AI by the bay conference which will take place in Oakland in November and I'm happy to be here with bar bar please introduce yourself hello yes so I'm bar sadurski developer Advocate at large both serial submitter abstruct and talks to different conferences and serial speaker as a result of although with not 100% acceptance rate and um organizer of some number of events have been on both submitting side and reviewing side of the cfp process okay then uh alexic off hi guys I'm Alexi so I'm the founder now Co organizer of you know scale by the bay data by the bay all the by the bay series of events this year it's a by the bay again following the pioneering 2017 version of it and uh I'm also the co-founder of the AI Alliance which is I BM met and 100 other companies also the uh founder and chair of Open Source science at n focus and uh member of representative newj at uh L foundation for and data uh also running Bop the oldest longest running deest technical mop on in the world and uh I also submitted uh multiple talks I am the program chair for the the Bas I reviewed 10 years worth of proposals and uh super excited to to start the CP process which will lead to the best conference in the world on AI in November and I'm happy to be here with Lisa as well hi everybody uh my name is Lisa I'm the product manager at a company called Data sto uh we created an open source catalog called Apache gravitino I also run a Meetup here in the Bay Area relatively new but I think it's really popular now called the data for AI Meetup and that's where I know a lot of these really great folks as well uh I also work at the Linux Foundation I'm on the LF and data Outreach committee as well as I run the open platform for Enterprise AI working uh group so for developer experience and lead the continuous St um delivery foundation's uh data Ops initiative so a lot of really cool things I also do a lot of speaking I've served on other program committees uh including the open source Summit North America in their data in AI as well as the continues still ret track so happy to be here so thank you everyone for joining me today with this variety of experiences and uh real deep expertise I think we're going to uh give people quite a lot of information on how to be successful in submitting uh talks because um cover is over now everyone is active I really I'm really happy to hear about this um data in AI Meetup that Lia you started and that basically one of the most popular ones in the topic in the Bay Area that's that that's really great um so I want to start our conversation uh with trying to shape what makes a strong conference um proposal so go ahead share your opinion only pick the speaker okay bar start all right yeah so um strong good convin talk idea makes strong proposal uh this is what it should really start with that you really have an idea for a good talk um expressing it in a good proposal is very important and we'll get to it in a second what to do what not to do but first and and and foremost you actually have need to have a good idea what you want to talk about good idea both in knowing what you are talking about and also something that is relevant interesting and and attractive to the audience and I would say getting those ideas um for technical people should be rather straightforward just tell us about what are you excited about and probably audience will be excited about that as well so if we're are talking about software Engineers it might be something that they work on they something that they learn about and want to learn more about uh it can be something that inspired them to dig into a specific content in the end of the day ideas that will eventually be the core of the talk is what drives a good proposal as well so I agree with your point but I was thinking about how I explained someone what makes a strong proposal and I thought that I would give them probably uh more formal uh explanation you know step by step like when I look at something what what the uh regular pattern what it's typically included in the proposal so when I came up with a list of things like I need to in my proposal I need to explain what am I talking about in general give a context like uh we're in the era of llm whatever and this is was too generic but basically I'm I'm trying to define the topic and then I move into to the talking about the problem like your why others should bother because you I found something that is relevant to them because they have the same problem or it might not be the problem but they might not even know that they have this problem but I need to explain them hey we do that absolutely we will get into all that in a second the structure bury delete don't bury delete be open about it try to make it intriguing there are tons of way just sping up I'm just spitting up but first ideas this is what we start with but hey let's let someone else yeah so a little formal as I said and then I present a solution and I talk about the concrete uh takeaway so if I can uh uh unpack my idea with this thing I think that's that's a really great start Lisa do you want to yeah can chime in here I think that O and B like you guys came up with some really awesome ideas and I think that the structure is super important right having a very tight opening like in the first sentence of reading it I should know exactly what this talk is about and what maybe the value is to me as a somebody in the audience I think one of the things that a lot of people get wrong in that particular first step is really scoping their talk correctly is this a wide talk or is this a narrow talk and how deep into the Tech um or the topics this talk is going to go into and often times and unfortunately this is something that we see a lot with say AI generated um abstracts is that there's a a very you know wide and Abstract scope and as a you know reader or as somebody on the committee I don't necessarily understand but what exactly are you going to cover what exactly are you going to say it should really read in my opinion like a thesis statement you know uh you know X because of Y results in Z and I should be able to fit the pieces there as a reader and as an audience member to what those benefits are and then you can list to me maybe some of the benefits or some of the things that you're covering um don't be scared to go into something that is unique uh especially you know if it's something that's a piece of technology that you know really well but also you know really make sure that you're communicating what those benefits are if I'm somebody who's coming in from a vertical or some other field Alexi do you want to add something absolutely so uh I mean everything Lisa and bar said is on the money I think the technical motivation is coming through you know there is this famous saying I think by one of the US courts like we cannot Define pornography but we know it when we see it same thing applies to intelligence it's very hard to actually Define what a good talk is or what a good speaker is but you know when you see them and in my mind in an engineering context at the meet up right you have a lot of very skeptical people you have a lot of like you know angry middleaged guys who will you know cut the crap and ask like what is actually going on here right will without any reservation and so you know that if you know people like uh you know pay attention ask questions that they they're engag so what usually I've noticed right and I think we've been very good at it I think for the 10 years and Lisa has been amazing at it like speakers are these kind of people right so they bring I think a lot of experience lived experience the not marketers could heard something com buy a bunch of BZ wordss together they suffer they show the SCS they still excited about it like they you know and they they share something that they've done and so it doesn't have to be fully open source right like we have all kinds of people talking about proprietary stuff but in my experience if you want it's and it's totally fine as well to Market your company to Market your product but you have to do it in a balance and so what I think you have to in order to get to earn the right to to sell you know some stuff you have to teach people you have to give right like you take people's time and attention to do something you have to give them back that is learning and you share some of the experiences that you live through you took some open source projects you put them together you got stuff to run you built a system you to it to customers right and and I mean even a life story right like I I'm kind of I think a lot of different talks have very interesting kind of things and so it's very minimal you have to tell a story the story is what holds it together and the good story comes from lived experience or some really kind of you know deep thoughts that you processed yourself right and it's kind of and I think this is how it's different from my because you know we need some human story we need the speaker to come through as a human being who lived through something interesting and got a story to tell and you will know that as opposed to somebody who just like droning on against some uh points and so at in by the bay events we had a very similar ask usually which affects you know the CP show some code right like and the code is a very huge differentiator because if you cannot show any code and you can just wave your hands that's easy but the moment you have to put and the code is an artifact you can bring a roboty Bild show us something you've done the moment you do that right you open yourself up to scrutiny and people can raise hand like what is this it doesn't work like right there is a type in there so so you immediately put people on the spot so by willing to do this you basically cut off a whole bunch of people who don't want to do this right like it should be falsifiable you know the the like the The Experience should be PR to questions and and they will see what's real what's not that's a nice one I I thought about like when when when you see a good uh speaker you know that this is one so when you see a good proposal you know this is one maybe uh folks you could uh give me an example like who's your favorite see see like a speaker who writes um awesome uh cfps in examples it's kind of hard to pinpoint like speaker that their cfps are way above others um let's let's talk about different styles of cfps and different group styles of cfps right so and this is something that I kind of wanted to throw out there and see what other think um there are I would say two main styles of cfp submissions the one is basically show me the money right when you put what you want to talk about in the title very clearly and then in the Sol in the in the abstract itself you lay out both the problem that you are going to speak about how you're going to solve it what this stock is going to show very practical very out there very down to earth and this has huge benefits both for the cfp committee that know exactly what what they going to get as as long as the talk goes but also for the audience they will read this and they will um realize what this talk is about another style is more of um like pulling out an Intrigue so you might put a question in the title and you might Allure to the problem um in the body of the abstract but not necessarily giveaways and the the solution maybe uh kind of imply what you going to talk about now this obviously um feels more elaborate more interesting but not necessarily that's the right way to go because then CF the the cfp committee will have much harder time understanding if your talk going to have a merit and also when the audience reads it they might take it um in two different directions they might say oh I'm trigged I want to go and come um I want to I I want to go to this talk and see what they are talking about or they might say well that's too vague I would prefer another talk which has a very clear abstract so I would say the Intriguing abstract is a risk that you might want to take if you have um The credibility if people know you and trust you both in the community uh of the both on the committee and in the community and you know they will go I see a talk from OE I'm going to go no matter what because it's going to be amazing and then you can try an experiment with that but if you're just getting started if you want to increase the chances that your talk is going to be accepted be very specific both in the title and in the abstract tell people what you're going to talk about and it feels like hey but I'm giving the lead away I gave all the secrets in the abstract don't worry about it if it's interesting enough people will come to hear more if all your idea completely uh fits into an abstract uh text field probably you don't have enough to deliver a talk anyway well I I you know as a conference organizer and I Hope Alexa can back me up I I actually hate when there is intrig because like that makes me I have to go and check who that person is if for some reason I don't know who's that who the submitter is so I have to check and I'm like do they really know what they're talking about or is it just like a elevator PCT of something something marketing marketing whatever whatever um so not not not on the other hand if it's uh if all talks are straight forward well it depends on the conference for AI by the Bas since we're positioning it as a like really for practitioner for practitioners um then uh probably it would be easy for us if if it was all um very direct approach on the other hand we're also looking for innovator for new ideas for something to you know God forbid revolutionize yeah no I'm not I'm not saying that you cannot have new ideas fresh ideas maybe not finished maybe not the ones that you can just put out there and it's very understandable for everybody what I'm saying is don't try to put too much in trig and too much smoking mirrors in your abstract because afraid that people won't come if you build too much to too upfront about what you're going to talk about it's not a real um um it's not it's not a real problem it's not something that you really need to uh care about be very open and very deliberate about what you're going to talk about um here's the question so um name your bar um I'll start with you name your own since you gave gazillion talks your own uh title like your favorite one the title of the talk that you gave I'm a sucker for word play uh I know it's kind of cheap but uh but I enjoy it so when I see like a clever word play in the title I chule and kind of I remember I remember them better than other ones so for me the the this is where you can get clever and and do some fun stuff with the with the title again up to a limit of course it still need to convey what you want to say but a little bit of weat I would appreciate in the title so are you gonna tell us the tile though uh I don't have one like to out of my sleeve right now but you know what let me see if I can find something that I really like maybe you have your favorite title or something that you not necessarily yours something that you remember and it stick to goe and I think the one that is one of the more popular ones that has word play I think is like maintaining diverse maintainers and this is more of like a diversity talk and it's like how do you get you know folks in the open source Community to engage from all these different perspectives and I think that's probably the closest I have to word play I'm a bit of a a boring person I'm very straight forward with what my talk titles are I think probably the most popular talk that I have is just like open-source data Ops and mlops strategies and a part of that is because it's such a wide scope and so people want to know like what are the strategies if I'm looking to go into open source and I'm very good with really doing the landscape talk it's a very hard talk to give sometimes but I think that's part of the Intrigue of it is that there's a little bit of something for everybody in there and so generally speaking I think the strategy talks tend to do quite well these days compared to specific tooling talks because there's so many different types of tools out there and so people want to know how are you approaching this problem and why and what are you doing to solve it versus this particular tool is going to solve all of your problems I don't know what your problems are I just know that my tool is going to solve them and you know those type of talks can be very unsatisfying because I think as Engineers as Alexia has pointed out like we're very skeptical of sales talks we want to understand the design of something and we we just we know that the waves of tools comes and goes right we want to understand the expertise that you've gained over the years Alexa do you have your favorite type of uh I first of all I want to give a couple you know points about the mystery I mean the best talk is actually which looks very simple up front and then turns out to be mystery so I give in example so there are two speakers which I you know appreciate all the years so George hops the guy who broke jail broke the iPhone and then started to do self driving car company Kam AI so he was a speaker at the original AI b b and self driving cars we have you know OT the Uber cell driving company and and George and so George actually had a very simple talk in the definition right like talking about Kam however uh he just held the old in trans like he was this like Wonder K right like this wizard guy and he it was eight years ago he like literally look like a high school student with like a hacker with like a Woolen cap and he just had this style where he everything he kind of described was new and mysterious and so by the time he was half through the talk you could follow him anywhere like people didn't really know what's going on like it was a mystery he was folding the mystery of software here's the guy who cracked the iPhone here's the guy who have like 500 bucks you know self driving or computer which you can hook up to your Honda and suddenly you have so driving car so at some point I think he was kind of unfolding the mysteries of software and then at some point he he basically asked the audience now do you guys want to me to reveal the secret of everything and at this point like there's total silence and like people was just like waiting for him to reveal the secret of everything the most amazing thing eight years later I have no freaking idea what the secret was because probably it was not really a secret but the way he build it up right really uh made it uh super engaging so that's kind of one thing so I think a lot of really depends on the style and you kind of have an if you have an idea right what people so good speakers usually will will will be entertaining the other speaker I really like is Paka Nathan who is like a main state of data science and P community and graph Community he's been you know in the community for decades so uh recently learned you know from the Texas he was a Bookshop owner alongside ly vender at some point like the people who started the Texas right so like he's a guy who can tell a good story so he's now in graph community so he tells a story about Boston crime with drops right so and so he and I think that's a very good use case so if you have a really motivating use case use it because the question is how to solve crime with drops right like how can you not be interested in this and then he kind of goes into a lot of context and like the Panama papers and so forth and he shows kind of very interesting graphs and he gets kind of to technical points specific there but I think a very good use case motiva a talk is very important so what I personally like to do you know like NE J has a lot of amazing people using graph so somebody some customer motivating the problem like let's say General Motors put the whole car in the graph right like car is representing you know in gen digital twin so if you have people who really care who customers and then you have Engineers who explain how the technology actually solves the problem right I think this is a very powerful combination well segu a little bit from the um topic room speaking about a call it the weave so let's let's talk about the how Ai and the uh chat B and uh change the writing proposals so there are two aspects in um using and abusing uh llms with with the the first is very simple just the sheer amount of proposals that you can crank now without moving a finger and uh this creates a symmetry between people who submit proposals and people who read them because creating proposals with llm is very very easy now but to review them you actually need to spend time and read them and understand them maybe I can ask um the chatbot to review them and understand yeah no I don't I don't think we can trust them with reviewing the for now for now right so this asymmetry is very very hard on the reviewers and on the and on the conference organizers and if previously we wouldn't mind kind of welcoming all your ideas and it was pretty rare to see cfps that limit the amount of proposals today it's a norm uh and I think it's completely Justified because otherwise we're just drowning in AI generates and I think the AI by the bay is I would say a little bit extreme example limiting the number of proposals I changed it to three after you told me right but but I think anything between three and five is is completely reasonable and that kind of once we limit that and and kind of bring it out control that brings the other question should we write the proposals using AI so there are two ways to go about it the first is this you're normal hey this feels like cheating it's not you who write the proposal and and that's kind of a one way and then we need to batttle it we need to root it we need to find the proposals using tools that like um zero GPT and then ban them from ever being accepted very purist and very kind of um old way CRA like old crafted way of doing things the other one is and this is what I believe in the end of the day a proposal an abstract is a way to express what your presentation is going to do about and frankly I don't care how to express it record the Tik Tok dance that will express your proposal I really don't mind and if you think that CH PT can write a better abstract than you will maybe because you are not that great in English maybe because you're just not don't know how to express an abstract you know what go for it use the tools that you have to get the best proposal possible now now when you said that I have to reserve one spot for someone who's uh applying with Tik Tock dance like I have to do that just make a Tik Tok URL uh you know in the like a field yeah exactly right and but on the other side obviously we know that all those models have a certain style and this style is not necessarily what we are not necessarily helps us to understand the style of of the speaker it's like all this llm language uh that makes all the abstract look alike this is obviously not putting you as a submitter is in an advantage you are like everybody else like any other llm generated submission and you might get rejected not because whoever reviewing passionately hates llm submissions but just because you know what this is just boring this is like any other submission they all look alike I don't understand what they want to tell me and this is why I reject you not necessarily because it's written by llm maybe it's written by you but this is just boring and like everybody else so what I want to say if you can force llm to express your ideas in the way that are clear engaging expressing what you're going to talk about not boring different than others good go for it and you know what if you want to check if that's not different than others use uh zero GPT to actually make sure that you doesn't sound like llm even if you wrote it yourself and then you know what have Ed it in the end of the day if we have good coherent uh abstract of what you're going to talk about and it's appealing and it's fun to read I don't mind that it's written with llm perfect love it uh liisa do you have um what's your take on it so there was a lot of information thrown out just now and I thought that there was some really great insights in there I also want to say though that if you're a speaker and you want to give a presentation for 2030 minutes or even a lightning talk Etc we're going to trust you to have a certain basic level of communication and personality to be able to convey that information as well and when I see a chat GPT generated um abstract I worry about that person's General communication skills of if they can they deliver an engaging talk but also whether or not the rest of their presentation is also going to be generated right that's something that I'm really concerned about because the abstract typically speaking if you have all the ideas floating in your head it's really easy to to come up with an abstract even in just point form format and then you can feed it into you know a chat bot and it can convert it into sentence format clean up some of the grammar that's not hard uh but if you don't know what your talk is going to be about if you don't have the ideas already in place then that's when a lot of folks are going to lean towards you know llms to generate it from scratch they're going to be pulling the sources they're going to be pulling the facts and that's probably going to lead towards the rest of you know your your talk as well unfortunately and when I think of a great speaker I think of somebody who knows the the information so well like the back of their hand that they can just communicate and go on right when we think about some of the best talks and I'm kind of guilty of this as well because I have really wordy sides but the best speakers they just have like a random picture that they pulled from Google images and the rest of them is just talking right it's just them talking it's them telling the story they're not leaning too much on the external sources because it all kind of lives within them already all that information is there they know what they want to say they're very confident in that and so I I do feel like an abstract is a proxy towards so much of that so really if you're needing to generate something from scratch really think to yourself am I you know really passionate enough about this topic to be able to communicate it to be able to say it uh because this is your elevator pitch right this is and we are back to where bar start with uh an idea that is precedes the actual uh writing full circle yeah yeah full circle Alexi uh in our pre-show you said something interesting about the using AI um for for writing submissions yeah I'll be blond about it right so if you have nothing to say and you basically ask charp to amplify it you will have 10 times more nothing to say in elaborate form right which will be both boring and tedious and uh right and so uh again you know I think we're the last generation which grew up without uh these tools and so like my kids are using it just native right so like my son didn't know any Java he built a Minecraft server in Java using chbt without like he didn't bother way like I said like son I'll give you to our lecture about Java in JV and he's like no that I don't have the time now he has a server right so I don't know what it means but uh so the question is you know you're going to be using these tools and I can have a personal kind of example so I'm submitt to whole bunch of uh conferences and you know I have like uh ration high opinion of my own writing skills because you know I am a very good writer in Russia and I won the first Global competition for Russian online poetry 20 years ago and like it's kind of translates like I think you know have style and stuff like this and so now I got a very good manager you know my VP of the St Shin who told me just run this stuff from grammar and like and then I thought like you know like this is you know like why should I do this my is great so load my stuff in grammar and oh my God it finds all the stuff right like I'm missing the Articles I'm basically feel like I just got out the boat yesterday like all this stuff needs to be fixed and and what he told me is very interesting he said that uh you know obviously half of the people here are immigrants right s Francisco is not a city it's a cruise ship people get on and get off right uh the so so you know other foreigners like ourselves you know who write reasonably good technical English will will see through it but there is a whole bunch of people for instance for instance I know In some cultures to whom it's very important that your English is correct even if English is not their first language because they wrun everything through grammarly they pay attention to this so I think it's common courtesy to use some tools right maybe not use you know Char PT so you can use it judiciously right and I think you have to review it in my mind like I'm looking for Style I care about style and when you see some style uh you know that people are going to be special in some way right so that's my main problem the thing this your St may be lost but uh the tools can improve the quality of your submission and grammarly should not be beneath you or something like this right I think it will helps and uh you may not know this but somebody may be judging you right on some stuff like that so that that's kind of it's good to cover your basis but at least can assure you we're looking for individuality we not around through grammar but somebody else will be somewhere else yeah as a last note on that I think in the data Community we're a little bit dry when it comes to our submissions uh I think one of the conferences that does the best abstracts is probably Defcon to be honest if you're a security person every single Defcon talk that makes it in there because that's a very competitive conference it's always a compelling story it's always kind of fun there's a good energy to it and I'd love to to see that for our conference actually I I I want to WRA up this uh conversation today was talking about what uh we would expect for AI by the way um so I I I wrote like a long description of each of the tracks there are four tracks for the conference distributed system data F program and open source AI all of them are uh all of them are we're GNA focus on AI in this context and right now Bar Needs to Go so we'll say bye-bye to him and I'll continue thank you very much for having me that was a pleasure and uh people go submit to conferences AI by the way obviously but also others don't be shy be explicit your abstracts tell us what you're going to talk about and you will be just fine thank you byebye thank you um oh let me change how do I yeah um so uh there are four tracks and each of them we think about them in a context of AI like for example distributed systems since the AI is pushing the limits of how you build uh distributed systems since like you need um like some of the when you do like when you train new models they can't uh be uh like now they're so big you can't do that in one machine you need to put them in a cluster and then now it's even bigger you can't put this in one cluster in one region you have to do even this level of Distributing so we want to hear about things like that and um list of others then uh data so data infrastructure is very important for uh AI how do you move that data how do do you make it in real time because you know everyone is uh wants their uh information uh fresh not stale they want to see uh they want to understand the Insight um like they want uh people to be hooked on their website because it's so personalized and that only possible if you move data really fast if it's accessible across your plastic then the sful programming so uh by the way conferences are known for uh bringing for looking at the um software holistically like you understand the um everything so I already talked about infrastructuring data but then how do you write code like what tools do you use uh what makes you more productive and effective that's what we want to hear in this uh um in this track and then open source AI well that's self-explanatory kind of everything how it evolves uh you have a project and it's open source it's helping AI somehow we want to hear about it uh then some products that are you passionate about and that change the way you uh do AI pipelines and whatnot in your uh company fantastic share that with us or just ideas what would be great to have so that's that's really not quick but I tried to put that in few sentences and uh uh Lista like as the Meetup organizer you see where people what they're talking about right now what would be in your opinion uh great ideas for the topic for this four tracks I think one thing that we're seeing especially in the data infrastructure Community is how are we serving different types of data for AI and you know generative AI framework specifically right how are we able to manage our unstructured data or you know our data lakes or streaming data and there's a lot of convergence these days around consolidating infrastructure but also serving infrastructure in a way that's not just for traditional like analytical workloads as we know it so I think some of that is really interesting for folks these days is how do I optimize you know CFA for a like real time analytics or how do I you know leverage graph or how do I leverage my data governance infrastructure and so all of that I think is really fun Alexa do please talk about our tracks you're the founder of the conference you I want your soul here absolutely you know uh so phot social engineering is one of the foundational talks right and as you mentioned is kind of stepping back thinking holistically I think at this age we are in super important era of AI native programming right and everybody from now on will never program alone they will be always surrounded by copilot and uh we're very fortunate in the Bay Area to be in the middle of this I think the most of the folks building this tools are here so I follow def tools for very closely and so that thoughtful uh softare engineering track now now will affect actively be kind of thought programming with or without AI right so the art and craft of software engineering and I think this is crucial now because how we as humans control and ensure safety of the software that AI is writing and it's actually doing what we want will be top of mind right data I'm I'm super excited that you know Lisa is here because I'm a fan of gratina data Providence data security data governance to top of mind and not in abstruct shape right so like you know uh I think gravitino is putting your money where your mouth is like organize your data right reason of it so and I think it's very cool that you know it kind of now became not just data for AI right so it's it's data and AI together so that is uh uh super kind of uh timely uh distributed systems again we are uh known for I in top distributed systems I asked M to start the apach spk up with me together so he can not at multiple conferences I brought Jake RS to speak about Kafka when he was still at LinkedIn and obviously Apache Kafka is huge right I think I was one of the first folks to in invite a bun of fling people right and so huge fan of uh you know Dr like the the Big Data Systems scale they didn't go anywhere where they powering internet we just don't talk about it as much as before but they without them none of this AI stuff at scale would not work in our approach to AI has always been engineering which it's not that call when you think AI engineer we we've been talking about it for a decade effectively because you know whatever you do in your notebook is not going to run your bank right and how do you integrate whatever data science you have with uh production system usually written in Java and not Java is not going anywhere right how do you going to integrate these flows uh it's key right and so I think bringing the stuff to production in an established context uh is very important and it's missing in a whole bunch of conferences right because they kind of end with you know here's the model here's the chat board sure like where does your shb live how is it hosted how is API secured how is it scalable right how do you do devops on it right all of this stuff we are very well known for so we we have this whole context and I don't think you will get anywhere right you will either usually get software engineering conference or you will get science conference and I think we are the only ones who really do the whole the whole span of things and the idea that you go home and for your startup you know exactly what to do how to deploy it how to scale it now with I probably can just one person can do it all right like we are very practical in that sense we want you to take home all this stuff and use it right and and play with it right so with that I feel like we should stop here and this is a lot of information for people to digest and I think we're going to do more live stream and talk about different topics that bothers us and bothers our community and uh just for people who want to submit please do I am posting a link here it's session.com by-- Bay um and uh don't forget to get tickets as well on our web website AI by the bay .io uh register um we are selling early tickets so please this is the cheapest way to attend if you want to attend otherwise submit and be speak not just attendant B speaker yes so uh thank you so much Lisa for coming uh that was um incredible to hear your insights and I hope to uh see you in November or earlyer if um I fly down to SF all right yeah and I'll See You by the bay in November yes absolutely yeah right thank you