Leveraging AI and ML to Humanize Data Experiences

Amazing yeah, thank you so much catherine um hi everyone. Thank you so much for joining us at beacon this year and welcome to this session on humanizing the data experience. I am megan ohearn im a customer engineer with gcp and looker focused on data analytics um and today we will be discussing and really walking through the modern business intelligence platform.

And ill highlight some shifts that were seeing within gcp and really just across the entire data landscape, from this more traditional approach to business intelligence into something that we might call human first, analytic solutions well start by setting the stage with within the contemporary data landscape. So um, in what were referring to as the information age, um and ill highlight some trends that were seeing in modern day infrastructure, as ditto really becomes even more of a driving and central force within businesses from there. Well move into a retrospective of how

Weve arrived at these existing problems and solutions and then well take a look at where the future of analytics is headed and then to tie everything all together. Well showcase. What this looks like in the context of our gcp solutions with a live demonstration that illustrates and showcases how gcp and looker are humanizing information to more directly connect businesses to insight, so um lets go ahead and begin. Welcome to the information age so were all familiar. Were all living in this? This is a time when there is so much

Information available and so much data being collected, but were seeing this paradox so were witnessing companies drowning in information, but theyre actually often starving for real insight. So how has this phenomenon come about? The use of data has long played a key, crucial, very important role in running a business in todays world were starting to see that the sheer amount and the complexity of the data

Is actually becoming overwhelming and its becoming prohibitive, rather than enabling and empowering so this is a time when companies are collecting and processing terabytes of data data sets have become so huge, have become so complex and fast moving that were really seeing companies having trouble understanding and Actually, uncovering the insight within the data itself and were all experiencing this phenomenon.

Of information overload, and especially in data um, were just seeing that the amount of data that is being collected being processed being analyzed is actually making it more difficult for businesses to optimally leverage the data to make better decisions and to more effectively drive operations within their Business, so how did we get here and where do we go from here um? How do we actually tap into and and use more effectively the complexity and the sophistication of not only the data, but also the tools?

That are available for data processing exploration to make better decisions and start to derive real, tangible actual insight from that data okay so when we think, about business intelligence at its core we see this need and this, desire for data-driven decision-making and decision making is a key Capability for every organization, its, what keeps businesses moving forward and really effective decision making is what differentiates successful businesses from average counterparts and from their competition um. So traditional business intelligence was really the first iteration of tools that allowed organizations to inform their decisions with data

naturally this became incredibly valuable because companies and leaders could now use data to verify understand quantify decisions that were previously highly intuitively guided so the launch of business intelligence provided that direct visibility into current performance um allowed and served as a benchmark of how the company was currently performing and allowed businesses to understand the business impact of every single decision that they made um and so this value was apparent just immediately and that encouraged organizations to become data-driven.

so companies are shifting to this data-driven model and with that they would see and be able to quantify operational improvements simply by monitoring these key performance indicators um and in this first iteration so the initial business intelligence um platforms limited decision making and access to largely to technical teams right um so a lot of this was being handled by data experts who knew how to process and analyze the data and there’s this need and this a bit of a gap so from there were seeing this migration to the democratization of data and data.

access so were seeing this um companies are starting to provide access to decision makers throughout the organization across the organization not just necessarily limited to specific technical teams or data capabilities or data skill sets um data has already proven itself as this incredible valuable incredibly valuable commodity and so modern companies are starting to realize this and realize that insights need to be accessible.

everywhere within the organization so there’s this big trend and this big shift to have data actually embedded at every level and that is mirrored and led to advancements in these tools that could provide any user limited database skill sets little little data expertise um with direct access to get answers without needing any sort of technical resource and to work directly with the data themselves so at this point virtually every decision within an organization could now be backed by data.

and we start to see and continue to see data just permeate through the entire organization allowing decision makers really just at every level to become more confident in their decision making and we see businesses are able to make better decisions overall just by leveraging the data throughout so this is giving organizations the ability to truly become data driven in every sense and we see this take off right every company wants data every company wants to leverage data companies who are data-driven are seeing.

tangible concrete value coming from the ability to track monitor progress and to predict future outcomes and everyone wants it the generation becomes ubiquitous across industries across geographies and we see this global trend continue to really just take off while this new functionality and accessibility definitely provided unique value its also introducing gaps in the data comprehension and so its in and yeah and its introducing this um a bit of data chaos so if you can.

imagine if you have different departments operating separately making decisions separately theyre still using data sets but maybe theyre using different data sets and theyre defining the data slightly differently so even though everything is backed by data the data is slightly siloed is isolated or its actually limited in its ability to properly to be interpreted properly and to be leveraged by non-non-savvy non-technical non-data experts um and so keep in mind at this phase the amount of data.

that is being collected that is being processed its only growing and behind the scenes the data processing has become more is becoming more sophisticated and more robust and so by providing access to essentially everyone the data experts and the technical teams and leadership and decision making and decision makers really need to ensure that the data is being interpreted correctly and being leveraged correctly because of like when you think about a.

modern business intelligence solution we even just like one generation before the existing solutions before today those have these the sophistication and the ability to handle billions of rows of data to ingest data from multiple data sources and when you think about the complexity of the data itself that has that had with it and holds with it significant risk for human error if users and data consumers have limited data expertise.

and so companies now are finding themselves sifting through data really trying to decipher and differentiate between signal and noise and they come to find that data is really only as powerful as the insights that we can gather or the questions that were asking so we see the shift in focus toward insight and organizations really ask themselves are the decision makers that we are.

empowering with this data asking the right questions and in all of this information what is actually the key insight so one more phenomenon i want to bring to light that’s happening alongside the advances in data capabilities um so as ive mentioned before the amount of data is increasing the ability to make data-driven decision-making is data-driven just is widespread its a common best practice at this point.

to work with that were simultaneously watching the decision-making process actually become more complex because with this there’s this new newfound visibility behind each decision-making criteria and organization wide at this systemic level so were were watching and witnessing as decisions across and within an organization are becoming more connected more contextual and more continuous and so with that there’s a need for data to become just as centralized just as integrated and for the decision to be equipped with.

skills that allows them to navigate the complexity of the data especially this evolving decision making process so that’s where this next iteration of business intelligence is headed its a really a curated data experience that’s designed specifically for human consumption for humanizing information to really cut through the noise and provide a clear insight to the data consumer so where do we go from here what does this.

look like illustrated we are now at a phase where its no longer about the access to information right in fact weve kind of walked through and proven we have access to too much information so now the shift is to build tools that are catered to the human experience and optimized to deliver direct insight and at the forefront of this we have questions shaping the future like how do we put analytics to work for people who are not data experts and how do we leverage data to make peoples lives easier.

and to provide direct insight that goes beyond the questions theyre asking and what they think they need so when we think about the next phase of analytics were were seeing and yeah were expecting this drastic change in the analytics experience so um i think the key the core elements are making data more accessible when and where users need it so redesigning every step of the data journey to focus on the human data consumer and make that data consumption and.

immersive experience that’s woven through existing operational processes and workflows because when you think about the the traditional approach to business intelligence today where consumer a data consumer comes to this centralized access point they ask a question they find an answer that’s not really its no longer enough for to meet the needs of an innovative business and keep keep them moving forward bi workflows are changing theyre shifting and.

we need to allow for data to be accessible and leverageable when its needed and where its needed so this next generation of tools data will be embedded and integrated through all of workflows it will be data the data itself will be transformed synthesized modeled and operationalized so that it becomes actionable so rather than needing to decipher meaning behind a metric and search for an insight as a data consumer.

the the data will be able to the data will be delivered in a way such that this future data consumer will be able to leverage the power of the insight immediately and directly through a strategic recommendation an ai powered predictive forecast and something that is directly actionable were seeing so i think the next phase will integrate elements of ai into every step of the analytics journey to really guide users to discover keyers that they didnt even know they needed.

were already seeing traditional business intelligence being augmented with ai and machine learning throughout the entire data process and these will only continue to be integrated more tightly and more closely um and then with advances in user experience ui design the next generation will integrate data into existing workflows and interface the data directly into existing tools so that everyone is leveraging the data from platforms theyre already using and the data is just enhancing the.

existing business processes so what does that look like and what does what does this actually involve ill showcase live in just a moment but want to walk through kind of what youll see in that demonstration um a few core trends and principles so the first ive highlighted um a focus on user experience and ui in general so were seeing were seeing this everywhere you think about just even on on your phone um were interacting with techno tech technological interfaces constantly and were the experience that our interaction with that technology is becoming.

more optimized more human focused more cater to the human experience and analytics is no different the second piece is automation across devices right so again to this theme of integrating and interfacing data into workflows not everyone is at their desk logging into a platform not everyone on their phone needs to find them when they need it and it needs to be in real time so it needs to be live it needs to be up to date especially when we talk about it in.

the context of becoming operationalized and really actionable um and then ai machine learning i touch on this but really leveraging the power of learning and technology to provide recommendations to data consumers so those who might not be as familiar with statistical analysis or data science workflows how can we still this information into insight and provide a strategic recommendation that’s conceptualized to the actual information this person needs their job better um and then as i mentioned earlier were really just going beyond the traditional.

intelligence workflow to offer insight and drive intelligence at every phase of a workers journey um so with that i will go ahead and hop in and we can see what this looks like in action so lets walk through this integration of google clouds database architecture reporting tool to see how this works in this operational context um and so in this case we are working with a fictitious um freight tracking and logistics company and well walk through this use case of really this control panel this dashboard reporting that is optimized for.

the tactical manager and then moving into the different personas will be leveraging the power of this data um into someone more strategic and then ultimately into deep fiber data discoverer um and so yeah just to start so were logging into the platform we have this high level visibility of the entire fleet system so we have all of the active drivers all the active trailers their routes the associated revenue so we can monitor this entire fleet in real time and then on the back end whats happening is we.

have this geospatial data of each driver we have the sensors on embedded into the trucks um and so we have that full tactical and operationalized dashboard with this multi-dimensional visualization and view that allows us to really gather this high-level website if we want to maybe probe into one specific route so at this level we have this high level overview of this macro perspective of the system-wide all of the roots and the revenue to optimize our fleet paths um but if we.

want to maybe fill in deeper to one specific route maybe i want to look in colorado and i want to look at one specific route look at the revenue associated with that maybe i want to look at the drivers and get it have their specific intel vehicle detail i have all of that visibility and i can kind of navigate through this tactically and get the information that i need to perform my management of this fleet so behind the scenes we have these senders monitoring each truck individually and im receiving these dynamic alert.

notifications based on any sort of changes or anomalies within that underlying data go to my first notification we can see we have this geospatial exact precise coordinates of this driver um and i can see that there’s some sort of vibration out of range so these key insights these anomalies these are being surfaced to the key stakeholders in this case management um of the fleet to provide visibility i dont need to search for the insight right i need the data to find me when i and then with that we have this real time visibility and so we can integrate.

this um we can feed them into this tensorflow model leverage best in class ai machine learning and with that we have this recommendation delivered contextualized to me as the fleet manager um and this is integrated with that communication channel so we have this critical mechanical failure i can now take action i can initiate this swap i can send this this will integrate with that existing text message communication channel um this is my phone number if i go ahead and run this and then i check my phone lets see.

um and yeah so i actually just received the message to the driver with your vehicle has been detected with the recommendation of where to proceed so based on my geo coordinates the back end um recommendation engine we can send that optimized route directly to truck driver if you envision this whole scenario its a really its a full circle operationalized data driven process right and so that’s the vision its integrated at every step and really allows me allows our fleet to.

to be optimized based on what we think and and what we know is likely to occur based on the predictive analytics and the data driving that behind it so beyond just this operational and tactical management lets um now shift into more of the business strategy focus so if i imagine i am running the business and now i want to unders how that new updated inventory is.

going to impact our financial forecasts i have a full visibility into this is similarly based on that change what is now the strategic business impact of this new updated um inventory forecast when you think about each persona and having this holistic management system catered specifically to each stakeholder and their needs with the ability to um integrate this into their roles and to their existing workflows that’s where the power that’s where the platform becomes incredibly powerful um and so with that if we want.

an even more strategic a deeper perspective we can dive into our learning environment so this will give us that contextualized high level business overview so for an even more perspective we can dive into this high level overview dashboard and this is going to offer that contextualized visualization for the data-driven the more strategic data-driven decision-making so i have these high-level kpis that have been identified by the business um backed by our mission right so fuel consumption um we can understand how that helps in.

our fuel um maintenance getting deeper into how that is contributing to expenses and ut and then driver safety and the way this is this system-wide overview allows us to then ask questions like how is driver burnout contributing to overall revenue or um our trip incidents contributing and we can start to think about in this case our three core tenants maintenance safety fuel consumption in the context of these larger um kpis that we are committed to driving forward um and committed to maybe achieving a certain level of um and so again that offers really that full.

ability to contextualize this and to put this in the mix of the overall system and to make more continuous and more integrated decisions because that data has all been centralized and then because were leveraging look around the back end to drive all of this insight we can provide that democratized access again to any user who needs a bit more beyond what weve surfaced in either that.

initial control panel or that reporting dashboard were equipping and within the organization to ask any question and then derive that answer directly from the data so if we if we want to look at our different trucks and look at fuel consumption i can bring in the metrics that are relevant to the question that im asking and then within a few clicks i have a direct direct visibility into how that’s trending across all of our different options um so that becomes again super powerful i can ask any question a few clicks answers become evident.

and so in this use i think we walked through how that intelligent dashboard really translates and contextualizes the data allows for not only the comprehensive visualization but then the strategic recommendations the predictive modeling um and the ability to really operationalize that and integrate that into every workflow for every percent within the existing organization.