From Automation to Autonomy: Understanding Agentic AI

Drawing from personal experience, historical perspective, and a deep understanding of enterprise data strategy, Mehmet Orun demystified what agentic AI really means and why flashy demos often distract from the real work—building trustworthy, outcome-driven systems. 

His insights are especially timely for organizations eager to evolve from automation to autonomy without losing sight of the human element. 

Key Takeaways:

1. Not All “AI” Is Truly AI—And That’s Okay: Orun made it clear that many so-called AI solutions are actually basic automation or logic dressed up with buzzwords. The real value isn’t in the label but in whether a solution helps deliver repeatable, data-driven outcomes that serve the business.
   
2. Data Readiness Is Non-Negotiable: Before any AI initiative can succeed, organizations must ask: Do we have the data we need, and can we trust it? AI without context leads to hallucinations, poor decisions, and wasted resources. Data must be current, consistent, and contextual—especially when dealing with personalized AI or customer-facing agents.
   
3. Use Case Selection Starts with the Outcome: Instead of chasing AI hype, Orun encouraged attendees to start with clearly defined business outcomes. Whether optimizing customer service or personalizing outreach, identifying who you’re trying to help and what success looks like should guide the entire design process.
   
4. Human-Centered AI Is the Only Sustainable Path: Agentic AI isn’t about replacing humans—it’s about empowering them. That means designing AI interactions that support employees, protect customer trust, and secure executive buy-in. The most powerful systems balance automation with empathy.

For the session um I love stories stories is what connects us and one of the things I do when I’ve been preparing for presentations lately is I try to avoid spelling grammar mistakes as much as possible they’re a little awkward right so I finish my deck i upload it to my ChatGPT instance and say did I miss anything right are there any obvious errors mistakes in this document please review and give me highlights slide by slide rough information of my prompt so as speakers when we got our decks we were given images of you know recognized people instead of our own uh interestingly enough uh ChatGPT believes I look like this and it did not say that is the picture of Dwayne Johnson not Mtorin or it believes that I can someday exercise and look this way moral of the story we still need to embrace AI with a level of caution human interaction matters but uh this is typically what I would look like wearing a fedora or a cap if we don’t want to talk about data you want to talk about dancing dickens zapper styles I’m here for you

What we’re going to talk about in the next 35 to 40 minutes is a historical perspective on allure and complexities of AI then I’m going to provide you with some real perspectives real steps you can take away on how you can plan your own AI initiative the nice thing about this session is it actually lines up very nicely with Peters as well as some of the audience questions then we are going to talk about some high level criteria to ensure your agents are successful this is not a session I’m going to dive deep into what agentic is there is plenty of material out there rather how to help AI including a pink AI be successful and then I want to leave 10 minutes for us to have a real life conversation with your questions i’ve tried and experimented with a lot there’s more I don’t know that I do know i hope this will be a good use of your time now to make this personal however I do need to know a little bit about you so we’re going to do some post-launch exercise please your raise your hand if you’re primarily associate your identity as a business person now raise your hand if you associate yourself as a technology person raise your hand if you raise your hand twice cool um please raise your hand if you consider yourself an executive who’s responsible for outcomes versus the doing please raise your hand if you are primarily doing versus the outcomes it is after lunch we’re going to get to blood pumping the last question u this is a technology agnostic event but I can’t dive into different levels of that how many of you work with Salesforce or Salesforce agnostic technologies how many of you are looking for generic practices okay this should speak to all of you thank you for that let’s get started i want to start by sharing something we actually built so part of my responsibility is I’m the general manager and data strategist for of a Salesforce partner which means I need to worry about sales pipeline reliability where we spent money right things that many of your organizations worry about and one of the things we often wonder is how good is the probability in the probability field often times this is automatically calculated people don’t even bother changing it and often times people overindex on the accuracy of this now I don’t know if you can see it behind the screen or in the back it says probability of close is 75% what if there was a datadriven way to say pro opportunity risk exposure was 100% indicating I don’t think you’re going to close this amount by the state could this be helpful maybe i’m hearing maybe well who’s the person that said maybe and why would it not be helpful potentially one you’re failing to look at a number of different factors that would influence that think in terms of the dollar amount and timeliness so if you set the dollar amount low enough you potentially increase your probability of close but you impact your profitability in your cost of sale at the same time I could then lengthen the sales cycle and extend it for 3 years which potentially allows me to close but not in a reasonable time frame so without understanding the contextual understanding of all the data associated with it it’s meaningless to know if it’s going to close or not plus at the same time what are the various factors that lead to a close without understanding what they are you don’t know what the leverage you can pull to change things i hope the camera got that did everyone hear it um this is one of the reasons by the way I am not a fan of the regular probability field or other bits either because without context a number is not going to be as meaningful but one of the aspects is we can build different levels of detail that could feed into a probability without overindexing thank you for such a comprehensive answer and I do agree with every point you made and if I didn’t I would let you know too so one of the things this example comes from is we looked at what did we learn both as humans and based on what our data was telling us what was cause what was common about opportunity dates being pushed out consistently and what we identified was if certain things were not happening the opportunity date was likely to get pushed out for example if we did not have an NDA in place it didn’t matter when the opportunity date was if it was a late stage opportunity it was probably not going to close if next steps were not captured well maybe it would close maybe someone has the NDA in their email but we couldn’t know as an organization so we started embracing the idea that by analyzing data tying it to specific contextual insights based on the life cycle of an opportunity by looking at what’s happening at an individual record we could be smarter about whether this automatically populated drop-own value was indicative but what matter wasn’t a percentage what mattered was being able to tell per opportunity what critical information was missing that would prevent it from getting to the next level one of the reasons 

I like this example and that was the best answer I heard by the way when I post the question is we have to understand the business context in any decision we are making so I didn’t use the word AI in that entire example we are going to build into how to deliver such solutions because our goal is the outcome let’s look at a bit of a historical perspective on how technology has been evolving over the past 30 years uh any anyone else that was working with technology in the ’90s or ‘ 80s a few of us in the room right this is when we were moving from mainframes to client server to end tier we had purpose specific technologies we had ETL engines we were selecting what database we should store our data in you know message oriented middleware was the new kit and static internet pages with handcoded web forms was what was popular jump into 2000s entr architecture and platform as a service is the norm cloud computing is booming uh instead of coding what a common customer dimension in a data warehouse looks like we now have master data management engines and we believe closely it turns out there’s something called the golden record that always consistently completely understands a customer data federation starts coming up to try to retrieve data real time from associated solutions and we are embracing document management systems more and more go forward 2010 cloud solutions are no longer argued over what if the data is secure or not but companies like Salesforce have grown they’ve acquired companies that gets them into the B2C space also integration tools are unifying whereas few years past you would be buying a batch integration tool a real-time integration tool a messaging integration tool companies like Informatica are saying buy the integration tool data science is hot it is one of the best paid jobs according to many articles written data links are hot because people got tired of waiting for it providing liable data every wants to create their own single source of truth in the meantime in content management one of the cool things is you are dynamically creating content based on the latest version of the logo image tiny the product catalogs so the amount of reliable content generation starts picking

up until recently when we are look up to 2023 now with co there’s a huge push towards digital transformation people are trying to put more and more online cost and complexity starts becoming a concern because profitability is questioned call for data unification that many many many single sources of truth leads to solutions such as Salesforce data cloud mdn vendors are trying to beat CDP. 

CDP vendors are trying to be non-marketing solutions in the meantime chatbots are everywhere supporting omni channel solutions and then of course in 2014 OpenAI releases uh ChatGPT peter talked about this much more elegantly this is a massive disruption because everyone starts interacting with something this is the equivalent of e-commerce technology raising consumer expectations in enterprise solutions it’s what is happening in the ability to work with and take advantage of information the worst thing that’s happening also at the same time which was already happening for the past 10 years the ease and ability to create bad information intentionally or unintentionally skyrockets it’s not about how do you know whether what you get from CHP is accurate or not it is what you read on the internet what is never for so data reliability is a massive concern so for all of you that are accountable for outcomes or responsible for making recommendations the challenge we have is everyone is trying to raise the efficiency because the economy is depressed we are worried about falling behind because everyone is talking about we need to embrace AI but technology is changing so quickly that we don’t know what technologies we should use whether we can trust information whether we are making the right investment decisions this is what all of us are in this session for and I’m going to share with you my set of perspectives on this because history teaches us some patterns that I believe will repeat themselves and they are already starting to beat if I was to say the goal of aic AI is to enable greater efficiency would anyone disagree with that generalpremise what about goal of predictive AI is to also increase efficiency by finding patterns faster same is the goal of automation workflows

what if AI is not always AI or AI stood for something a little bit different the way I’ve been thinking about this and I was talking to a good friend of mine who runs AI innovation for Salesforce he said they reviewed hundreds of AI agent ideas and a vast majority of them could have been achieved with simple automation or other AI techniques the good news is and if you go back on the history slide if you are a consultant or if you’re a customer that have built your organization’s IT structure in the past 3 to 50 years you have some of each of those technologies which contains data logic buried in metadata you can actually take advantage of so our challenge is not how do we implement agentic AI tomorrow but it is to understand what is the business use case so we can take advantage of all of these capabilities delivering trusted repeatable information based on the needs of the use case i want to pause here to see if there are any questions or reactions so but how do you identify something to start applying these technologies

in the morning keynote there was a question on how do you select a use case frankly other than throwing a dart you have two options you can go bottom up if you already have an idea having an idea means you have a business stakeholder that knows that there’s a need that they want to be able to automate and they are willing to come on the journey with you the other idea is you’re going to go top down you’re not going to try to solve everything but you want to be intentional in what you’re suggesting be used or not be used either way evaluating whether that use case is a good fit for whichever AI flavor the questions are going to be the same this is a slide from 2019 when we used to guide people on the way to implement customer 360 solutions this is going beyond MDM by understanding your journey with the customer and deciding where to start in 2015 I would submit to you the goal is actually exactly the same what you want to do is decide who you want to engage are you trying to optimize your recruiting your supplier relationships your customer relationships if it is customer relationship pick a customer type and draw the coarsest grain customer journey you can envision from awareness all the way to adoption growth or attrition the value of this is ultimately we are trying to ask a couple of key questions what are areas with highly repeatable tasks where we can easily demonstrate delays or inconsistencies will cost real money either by losing revenue delaying revenue or incurring unnecessary costs and then within that particular engagement point are there specific tasks or jobs we can actually efficiently streamline make sense so far any questions okay so once we do this then the next step is you pick a moment you describe the ideal outcome in that moment this is the equivalent of the output in Peter’s algorithm we didn’t coordinate but I was really excited while I was watching him you then identify which personas are influencing that moment because when you provide an output someone needs to see and act upon that output it needs to be clearly defined otherwise we are not going to be able to answer some of the better questions you determine what information they need to reach that outcome it may be a purchasing decision it may be saying the field service repair is now complete and the customer can continue using the machinery it may be a support case closing and then within that you determine repeatability of the task because if it is not highly repeatable it’s not a great case for automation if it is repeatable you also ask do we need to incorporate empathy as part of the decision making because empathy is still not something you can build into the algorithms

once you ask these questions you put them in a simple table never have just one if you have more than five to seven it’s going to be overwhelming you want to start small enough to repeat and then you come up with a table like this not knowing the audience I picked a few different examples from a few different scenarios so if you’re a nonprofit the first time you get a donation from someone or you think it’s the first time you get a donation from someone you want to send my thank you automation can help you determine is this the first time you’ve seen them because the journey you want to put them on matters if it’s a first time engagement or they donate donate annually I love United i’ve flown with them for couple of million miles but every time I renew my 1K status I get a congratulations on becoming a 1K i’m like can’t you just see if I was a 1K at least last year and say congratulations for renewing your 1K that would make me feel a bit more understood right real example but your message you sent can then be generative about what you know about the person if they responds how you mentioned response can be agentic and this is by the way probably three phases of a project where what we want to do is we want to nurture that relation into a longerterm domain what I highlighted is the moment we would normally encompass this under a broader use case now if we’re doing marketing segmentation I want to define what markets are effective i’m probably starting with some level of analytics some level of predictive intelligence i want to see my patterns quality of the training data set is going to be important and once I figure this out I can see how I can apply further automation downstream but at a super high level the way I’m approaching marketing segmentation dealing with massive amounts of data is very different than I’m engaging with one record at a time that’s transaction hopefully that’s making and then 

let’s imagine a retail scenario raise your hand if you ever bought something online and then keep your hand up If something happened to that order that you didn’t want keep your hand up if you try to reach out to customer service and did not have a great experience awesome so this is one of the most easy to break down and explain pieces because when you can play pretend and think about what were those moments that could have been improved every single one of those moments is a micro project phase that you can put in place by identifying the outcome you want to achieve the information you need to be able to guide to that outcome and build a solution so you can start with predictive there is a storm and a bunch of shipments are going to be delayed send an email alert saying do you want to cancel if it’s not perishable or you know how do you want to get an update depending on your response if someone is reaching out to customer service recognize that your customer service automation and your delayed order follow-up flows may need to be related to each other so this is one of the powers of looking at your processes or customer journeys end to end while being able to execute moment by moment make sense hopefully any questions yes sir i would put a more nuanced view on it okay hello i’ll be loud enough okay if you’re talking about what you’re talking about you’re talking a very broad consumer and you’re treating everybody equally you really want to look at it from a a a genetic perspective why not be very highly personalized you provide a higher quality of service to very specific targeted individuals or such that much have a higher lifetime value and potentially repeat and if it’s somebody new maybe you offer it because it’s a new customer but if someone’s a lowv value customer why do you give a crap it maybe cost you more to serve and to build the implementation to support them so I think you look at the data of who your customers are understanding your customers look at that and then build a bell curve associated with that and look where you’re going to target your resources to be able to solve and you’ll have a much higher repeat business for people that far more profitable particularly if you’re looking at profitability of the products that they buy uh were everyone able to hear the answer now I’m going to add an amend to it it’s not just what they spend with you it is their propensity to buy based on information you may get well it’s not just lifetime value because lifetime value is past transactional habits it’s also what you know about the demographic and if you want to foster the relationship with that demographic to not lose potentially large customers in the future this is the reason we need to look at these as broader solution architecture by the way and many of the slides on the STE we can probably do distinct talks also i will touch upon that specific example h about 15 slides yeah lots of slides so the thing to remember here each idea can be broken down into individual distinct business valuable

steps so in terms of planning your AI initiative no matter what you’re going to do a key question you need to answer is do I have the data I need for this initiative to be successful and if I don’t can I produce it so um I want to tell a story and this is a story with my father who turned 90 recently he’s a retired frig here’s a military engineer he is used to being given a mission and you happen to discover what tools you have to get the mission done and part of his gift um when he turned his turned 90 was he wrote a book of his memories which I didn’t realize until after I put the book cover in this image how relatable it is to what we were talking about there’s an old man looking at a store carrying a baby with the sunrise or sunset we don’t know the time of day because we don’t know if this is the end or beginning of a new era but he was asking me so explain to me what this AI thing is so we interacted with Chachi Piti and I gave it the prompt uh describe my father’s name a retired soldier didn’t provide any more details branch rank etc and we got an answer on what uh his rank was when you’re a general you are probably a little better known in the public space where he went to the schools from a military academy that was it then I said what else you can share about him he’s written books on poetry after he retired he was involved in the industry so we got a little more details he’s like “Oh this is interesting where does it know?” Okay it makes sense like I know that I had written things then I said “What do you know about his son?” So these are the actual dramas and then Chia says “I don’t know who his son is.” So my dad goes “So why doesn’t you know that we are related?” And I said “Because government records on parent child relationships is not public domain so it’s not going to know just because we may have the same names that we are related to create that relationship and being a military engineer of how parts fit in a tank a gun a submarine in a bridge he’s like “Oh that makes perfect sense.” Which by the way was a 20-minut conversation drinking Kgnac that may be a three-month conversation with an organization they are still trying to understand why you may need a data platform to achieve enterprise AI solutions rather than why don’t we just see if we can build something as enterprise AI than build your own piece uh you will hear that I do not prefer build your own AI as a P&L leader by the way but a bit of a story for you so moral of the story data and the technology we are using to achieve outcomes matter so we’re going to talk a little bit about the data first and technology at the end of the day and um I should get to know the 007 uh laptop’s name because we’re going to engage more i’m sure Andy Andy also we will interact more this is one of the things people need to understand we can have AI technologies just on unstructured data right your typical ch you’re dumping your documents it’s doing internet research or dumping multime media is essentially that it is pretty cool it’s valuable you can have deterministic AI by throwing in customer or transactional data and of course if your customer data is disconnected you’re going to have issues enterprise AI means you know when unstructured data relates to a customer which knowledge article relates to products a customer might have purchased what you know about customers history when all of these things can be served in a trusted contextual secure environment so part of the journey as we think about the road map is understanding exactly where things go and assessing the risk and in most organizations data does not live with just one application even if it did they are distributed across records we need to unify this the good news is even though every organization I’ve ever encountered with my own included has some level of data challenges when you ensure the information you need for a specific business outcome is complete consistent current correct contextual and compliant enough for that use case it doesn’t matter what flavor of AI you’re thinking about you’re going to have the right foundation so this is not going to be an effort that is going to be wasted if you’re a data professional like I am this is one of the best times to deliver value to organizations across the initiatives in a rapid way so in my perspective there are five things we can identify that indicate a level of risk for AI hallucination in organizations number one is incomplete understanding of the customer i have now seen demos and keynotes that talk about do not send a customer an order when you have an open case and they are frustrated yet 15 years later we are still dealing with that exact same challenge if the records are disconnected there are confidable ways of assessing it your personalized AI is not going to be as personalized number two do you know when you have missing data it may be missing records it may be missing data in fields not every field not every record is as actionable as another once you determine what fields you can rely on and yes this is a structured data perspective but I believe in enterprise AI structured data matters for those personalized interactions you need to be mindful of certain metadata because this is where you’re classifying sensitivity and when you work with reasoning engines this is where reasoning engines are identifying where else they can go and fetch the data if you did not configure prompts explicitly with those fields uh two other items you know traditional information life cycle management challenges if you put out of state data whether structured or unstructured it’s going to skew your results and if you do not monitor your data trends and metadata for change the system may break and you never knew that there was a leak in the pipeline just going through this is normally a talk on its own i’m going to go through some of the slides quickly and invite you to connect with me if you ever want to geek out on any of the details um couple of stomachs I want to leave you with is when we are engaging with an individual because systems still do not send messages to companies they send messages to individuals we need to be clear who we are engaging it with and what are we supposed to know i often bring this example and say are these duplicate records now in a typical system there are probably like 50 records for me but if these were two and if you don’t have an opinion that’s fine raise your hand if you think they are duplicates raise your hand if you don’t think they are duplicates raise your hand if it

depends so for the last 20 plus years organizations have been obsessed with the idea that we need to eliminate duplicates and merge things as much as possible not realizing what’s more important is to be able to deliver contextual information to a distinct persona when and where they need it so if someone wants to engage with me based on what I’m doing at Pure Nova my Salesforce transactional history or unstructured activity history should not be a part of this other than perhaps knowing I came from Salesforce and it is an attribute of me as an individual if you’re on the technology side and you’re thinking about your data architecture we need to think about engagement with individuals versus engagement with business contacts and these are the type of solutions you’re going to need to think about the good news is as you analyze structured data content which is a lot easier to do still unstructured data technologies are fast evolving you can quickly see what is the data pattern how often do I see low level of uniqueness for example we think this is somebody’s personal business phone it’s only unique 40% of the time it probably means we inherited it from the company record you can come up with record designs by identifying what fields are no longer being used and then focusing on what fields are reliably populated recently you can narrow the scope for any analytics or AI technology to fields more likely to provide value of course there’s a contextual depth by looking at the data content behind the fields and identifying how many fields are always populated but they only seem to have one unique value only you can identify what you may eliminate and to give you a real number I was working with a customer uh advising them and they had 589 fields in one object alone and it turns out they could only use 73 fields reliably and we were able to identify in a number of hours applying wellestablished data profiling data management techniques so data content data readiness matters before we decide what to do the reason metadata matters is in enterprise solution salesforce agent force is a good example of this it’s going to use configure prompts it is going to try to guess what you’re trying to do based on metadata available so if you have a field called score with no description it’s not going to know what it is and it may either use it incorrectly or it may not use it which would also be a bad outcome so and then the last bit is you can quickly and easily see out of state data that you may choose to archive purge or just remove from or not feed into your um rag your cloud-based databases whatever technology you’re using not just for data reliability but we are in an increasing consumption based model today every bit of not necessary data you feed into a solution is costing you money while you are still not using it 

so for the sake of time let’s talk a bit about technology options i’m going to speed up because I want to make sure we have time for Q&A when you’re planning your AI initiative these are the five phases I am observing organizations going through what is dominant right now is what I’m calling standalone unstructured data for a chatbot or generative AI experience you are giving a bunch of data it’s not personalized you’re getting prompts they can be well written or poorly written valuable i use these it didn’t know I wasn’t going but still super efficient it is easy to get into low barrier to entry then there are solutions that are dealing with application metadata because it’s not data it’s not considered as sensitive this is helping people better configure better code better test but in of itself this does not generate tangible significant business value the fun journey is on the right side what can we do within a business application about a customer or a supplier or a vendor how can we do it even better when the data is unified in the application and even better when we combine it with unstructured data and data from other systems as well this is where we are all trying to move whether you’re a consultant or a company and this is why I think people come to sessions like this for the race my personal belief on technology is platform players are going to be the winner and I don’t consider AWS or GCP a platform player in this case by the way they are technology toolkit providers in the 80s we had batch tools data warehouses operational reports these were all standalone just building this quickly things started coming together integration tools unified things uh marketing focused CDPs brought master data management type capabilities ities for the benefits of the marketers modern analytics like Tableau started providing an NLP natural language interface to interpret results what we are talking about now is modern platforms are bringing in unstructured data databases identity resolution structured and unstructured data handling workflow real-time data lookup into the architecture and while anyone capability within may not be as good as something else I can buy and integrate today my perspective the rate of change happening in individual tools is so high that I am personally betting on the platform plays the way it looks right now is if we look at some of the examples we have traditional marketing CDPs in this space as well as the business application providers that enterprise solutions if you look at some of the key capabil ities you are seeing much of what we need to wire are being implemented to a degree larger companies are also by smaller companies so this is something to be mindful of in general in the way we’re looking at things I know I sped up my thinking am I still making sense in the days I’m going great uh by the way I spent a lot of time with Salesforce’s technology and not time with the others so if you want to ask me questions about it after break I’m happy to answer or during Q&A but this is a more comprehensive way of looking at it because it’s a non-commercial session so we talked about use case selection data readiness assessment and perspective on how you can go after technology let’s talk about the other factors that would make agents successful because agentic solutions working with automation distributed APIs etc is part of what we are doing today i’m going to focus on the example of support triage and customer service why because this is automation as well as agentic potential this may require empathy which means it cannot be fully automated but a lot of things can actually be streamlined including generative AI summarizing past case history including data unification to know what do we know about this customer their order history and then being able to route it to a human when it is called for these are four areas that you can look at salesforce actually has a free trail on this that I recommend it will take you 15 minutes to complete it is non-technical you need to identify where you want the agent to engage when that agent should hand off what are the conditions if there is routing logic from one channel to the other cuz it may start on the chat it may switch to phone you want to not lose any of the data otherwise you lose the efficiency benefits anyway and also what data it is supposed to know about and act on because the security rules for an agent is different than security rules for an employee or a customer you need to think of agents in a given test as if they’re a distinct role in your design for more details take the trail it goes into all of these four buckets

the thing that’s going to be most important for the companies that are getting ahead is going to be the human factor and I submit to you there are three types of humans I worry about when I think about rolling out AI customer is one because if they think the way we are trying to be efficient is actually at their expense it doesn’t matter that we are more efficient we are going to lose customers unless we are intentionally trying to meet some out to Andy’s point earlier executives trust acceptance and embrace is going to be important because no one has unlimited funds we are all writing checks based on what we believe would happen so we need to understand how any initiative ties to a tangible business value outcome and why we think this is the best alternative today that is going to be extensible into the future and for our employees which is all of us also by the way whose jobs are being impacted we need to uh support them with the type of not just technical skills but empathy training so we all understand where we are in the journey because it is disruptive at the end of the day though the key question is who are we empowering when why and how and how would we know if we are doing it well and successfully in conclusion to have time for Q&A allure of AI is real complexities are real what we need to do is learn from the past on how we can apply these into the future experimentation because it is so easy right now is something that you should embrace and get going but do not forget the human factors in technology funding before doing the Q&A I want to call out one thing that is off agenda um so I’m here then rushing back home for my birthday and every year I pick a nonprofit because giving back matters this is a time nonprofits are struggling and we can talk about all of the causes to help in all the problems in the world but early childhood education and learning to me is what matters the most when we want to give human beings the ability to be self-confident to learn to challenge ideas for critical thinking so I chose the Chicago based nonprofit called Start Early that operates nationally if you are willing to give back great if you are not able to give back taking care of yourself before helping others is a really really important factor i just wanted to mention this also cuz it’s a learning culture with that I got my 10-minute warning so we can have 10 minutes for Q&A and thank you for listening