Neurodiversity, Data Labeling and AI
On this episode of Not Your Father’s Data Center Podcast, CRO and host Raymond Hawkins, chats with Chief People Officer of Enabled Intelligence, Lauren Bacon Smith, about neurodiversity in tech, data labeling, and AI.
In 2022, nearly 21% of persons with a disability were employed, but were much less likely to be employed than a person with no disability. For one company, however, people with disabilities are an untapped talent pool with lots of potential. Growing their data labeling and AI business, Enabled Intelligence looks to build an inclusive environment for those with neurodivergence, as they see what others might not be able to.
The two discuss:
- How tapping into the talent pool of neurodivergent people led Enabled Intelligence to success
- How Enabled Intelligence is growing their core business
- How AI modeling helps label the dataset faster
“Yeah, so we actually have a data science team in-house now where we are actually creating some of our own models, so we are kind of having that full life-cycle where we have also done data curation, so clients coming to us, a lot of times have their data already, but we have the ability…to actually go out, get the data, prepare it, annotate it, and then we also are able to create and test our own models,” explained Bacon Smith. Lauren Bacon Smith holds over a decade of experience in program management, human resources, and recruitment. Starting her career in the hospitality and service industry, Bacon Smith spent eight years with Hilton Worldwide as the Senior Manager for Military Programs before joining Enabled Intelligence as the Chief People Officer to help continue fostering a culture of inclusion.
Read the full episode transcript below:
Raymond Hawkins: Welcome again to another edition of Not Your Father’s Data Centre. I think we’re like episode 54 or 55 now. I lose track. They’ll probably yell at me for naming the number of episodes, but we’re in the mid-50s. Today, we are joined by chief people officer, one of the coolest titles you can have in corporate America, Lauren Bacon Smith, who is with our friends at Enabled Intelligence. Lauren, thank you for joining us today.
Lauren Bacon Sm…: Thanks so much for having me. I’m excited to be here.
Raymond Hawkins: So I’m in Dallas, Texas as always. Not always, but almost always when I record. Lauren, where are you?
Lauren Bacon Sm…: I am in the Washington DC area, in Northern Virginia, Falls Church more specifically.
Raymond Hawkins: All right. All right. You can see the Capitol from there, can you?
Lauren Bacon Sm…: Yeah, probably over the tree line. Yeah, not too far away.
Raymond Hawkins: All right, very cool. Well, love getting to hang out in DC. Does frustrate me occasionally. We won’t get into politics. We’ll just focus on business here today. So give us, before we get into what Enabled Intelligence does, and how you guys serve your customers and our government, would love to hear a little bit about you. How did you get into business? How did it get started? Give us a little bit of background on you personally, and then we’ll transition into Enabled Intelligence.
Lauren Bacon Sm…: So prior to coming to Enabled Intelligence, I actually worked for Hilton Hotels for almost nine years. And I helped start up and run their military hiring programme there. And we really just kind of had a really big focus on culture and people and inclusion in general at Hilton. But my main focus was in that military space, so really kind of having a strategic specialised hiring and retention programme.
So when I had the opportunity to be introduced to Peter, and find out about Enabled Intelligence, and the company that he was looking to start up, I was really intrigued with the idea of making a switch from a 100 year old massive international company to a brand new startup that really didn’t exist yet. But the biggest pull for me was the people side of things. And the fact that Peter was looking at starting a technology company, but the first person he was looking to hire was a people focused position. So to me, that really told me that that was the type of organisation that I was interested in joining. So I’m Northern Virginia native, grew up in this area, so I’ve been here most of my whole life, and was kind of excited to do something totally different.
Raymond Hawkins: Gotcha. Yeah. What a switch, a 100 year old real estate/hospitality business to an idea. That’s a bit of a risk. What would love to hear a little bit of how you decided to take that plunge from a just personal risk perspective, probably pretty established, pretty stable trajectory of what you would look for in Hilton, go into what was nothing more than a concept at the time. That’s kind of crazy. It’s worked out. But still probably a little scary.
Lauren Bacon Sm…: Yeah, yeah. Well, I will be totally transparent that I did have the push of COVID, and the hospitality industry taking a massive hit. But people kept asking me, when I was looking for new opportunities, what industry do you want to work in? Which companies are you targeting? I was very established in the military space. So I think a lot of people were pushing me towards staying in that, the military realm. But really the biggest things, when I was sitting there trying to think about what it was that I wanted to do next, was that I wanted something that I had the ability to build and collaborate on without a bunch of red tape and a bunch of hierarchy, and having the ability to impact people. And being with other leaders that were on the same page about the impact that focusing on people has on your bottom line and on your business.
So I never had considered working for a startup, or even probably a company as small as Enabled Intelligence, even as now. But once I got introduced to Peter, and kind of learned about his vision and learned about his tenured career and successful career he had had. This is not a startup being run by a 23 year old recent grad in Silicon Valley. He has the experience to back this up. So learning about this opportunity was something that just really actually aligned with what it was that I was looking for in the opportunity, regardless of the type of company or industry that it was in. So I decided to go for it.
Raymond Hawkins: Very cool. Well, before we get too far down the path of getting into Enabled Intelligence, let me first say, as a marine vet myself, I want to say thank you for being concerned with helping my brothers and sisters in arms find out what’s next. It is a tough transition. And for our military that’s been engaged in armed conflict now for 20 years, I know we wrapped things up in Afghanistan, but that was tough. That was tough on the services. And it’s difficult to transition from military life to civilian life. It’s even more difficult to transition from combat to corporate life, civilian life. Civilian life that’s not involved in combat. And we’ve got a generation of our military that have been in combat for 20 years. And that transition’s hard. And I appreciate you being part of the ecosystem to help them make that transition and that you did that for so long. Appreciate that. As a vet myself and as the brother to two other Afghan and Iraqi war vets, it’s tough to watch them transition. It’s hard.
Lauren Bacon Sm…: Oh yeah, it is. It was an honour to work in that space and assist such deserving group of people. And definitely appreciate and grateful for you and your family’s service as well.
Raymond Hawkins: Yeah, I think people struggle to understand how much war impacts the psyche of our fighting forces, and how tough it is to transition back. And I think we look and see their struggles and don’t understand what’s underneath and how hard it is. So your heart for that is important to us, me and to us on this podcast and at Compass, like you mentioned briefly, you wanted to be focused on people. We’re a land development and data centre construction business, but we start with our people. We’re big focus on culture. We believe culture and people are how the business grows. And that’s important to us, and would love to transition and let you start to talk about Peter’s vision and what Enabled Intelligence, what made him have the idea, why do y’all go about it the way you do, what it is you’re doing. Let’s start there and we’ll build from that point.
Lauren Bacon Sm…: Yeah, sure. So I shared with you earlier, I am not a technical person. Obviously I’m on the people side of things. So the way that I explain what we do here at Enabled Intelligence is definitely more in plain terms. So I guess I’ll start there, just to set the stage for what it is that we do here at Enabled Intelligence. So we do data labelling and data annotation. Well, that’s what we started with, and still is the bulk of our business. But the example I always use is you know when you’re signing onto a website and you are asked to prove that you are not a robot, and most people use CAPTCHA as the tool to do that, and you’re given the images to select every-
Raymond Hawkins: Which ones have sidewalks? Which one has mountains?
Lauren Bacon Sm…: Motorcycles or whatever.
Raymond Hawkins: Yeah, right, exactly. I’ve seen it. Yep.
Lauren Bacon Sm…: Yeah. And they are these, to us, pretty blurry images that you’re looking at, but they’re all street level, eye view images. Someone has had to sit there and label thousands of images of crosswalks, or stop signs, or cars, or whatever the object is to be able to train the computer to be able to automatically identify those through AI technology. Those images are actually very clear compared to the type of images that we work on. Most of our client base is federal government. So we work a lot within DOD, a lot within the IC community. And we are working more on geospatial imagery, satellite imagery, so aerial imagery, where the objects that we are looking for are much, much harder to identify and getting into very specifics of not just a car or a plane, but to looking at, that’s a MiG-29. That’s a very specific type of ship. We’re looking for lots of different types of objects.
Most of what we do is imagery, but we work on full motion video. We work on text, handwriting projects, audio, kind of a wide variety of different types of data. So our annotators are going through and being trained to become experts in identifying these different objects or ontologies, and then going through and actually labelling these images to be sent back, or labelling the data to be sent back to our client. So they can use that towards creating an AI model to be able to have technology to automatically identify and alert for these different things that they’re looking for.
So when Peter came up with the idea for the company, and he came up with this idea years prior through working at different technology companies where this kept coming up as an issue where, to move forward with projects they were given, they needed their data labelled, and there wasn’t anyone to do it within the US because most of this work tends to be outsourced to overseas where we cannot send DOD and IC data and information. So the kind of concept for the actual technology piece started years ago.
Raymond Hawkins: You don’t want the Russians confirming that’s their MiG in the picture? Hey, can you let us know? Is that-
Lauren Bacon Sm…: They might be most accurate at that.
Raymond Hawkins: Looks like your MiG on the… They might be really good at it. Yes. That is a MiG-29.
Lauren Bacon Sm…: Yeah, exactly. Yeah. Here you go United States military.
Raymond Hawkins: We’re glad to tell you. I get it.
Lauren Bacon Sm…: Yeah. So we then started thinking about, okay, what is the workforce going to look like that we’re going to hire to do this? And Peter had heard about a programme that the Israeli defence forces had, where in Israel there is two years of mandatory military service for everyone. And one of the previous disqualifiers were people that were neurodiverse. There’s a lot of different things included in neurodiversity, but one of the more well known diagnosises within neurodiversity is being on the autism spectrum, just for one of the examples.
But they created a programme, two different programmes, but one of them was in their cybersecurity sector where they started aligning some of their neurodivergent population into this programme and started putting them into these programmes that they were just really, really excelling at. And some of the characteristics that they were listing off was extremely detail oriented, process oriented, being able to recognise patterns and analytical thinking. Not always, but sometimes more asocial and hyper-focused. So we started looking at some of the work that we were going to be doing and a lot of these things aligned.
In the US, this population, also unfortunately, is a lot of times overlooked. They are either unemployed or underemployed. So it’s also a readily available talent pool to pull from and kind of untapped talent pool to pull from. So we decided that we were going to have a mission to have a focus on hiring people with disabilities, and a big focus on that neurodiversity piece. So we set out with a goal to stay at at least 30% people with disabilities. Originally we had actually planned to have most of our team members with people with disabilities, but after consulting with some different partners that specialise in this space, we actually were advised to create integrated employment opportunities.
So just having it be a normal job opportunity and not a specialised programme like some other companies are out there doing. So, it’s not a different job opportunity, it’s not a different set of expectations, it’s not a different training programme. Everything is the same for everyone regardless of disability or not. But we have really built our company with all of this in mind on creating a very inclusive environment from our job requirements, the way that we do our screening process, to the way that we do training and communicate different things within the company. We have done all of this to try to set up our diverse workforce for success.
Raymond Hawkins: So Lauren, you said that there’s data labelling and data annotating is the core business. What other things have grown from there? What other things are you guys doing?
Lauren Bacon Sm…: Yeah. So we actually have a data science team in-house now where we are actually creating some of our own models. So we are having that full life cycle where we have also done data curation. So clients coming to us a lot of times have their data already, but we have the ability and have had clients that have come to us and asked us to actually go out, get the data, prepare it, annotate it, and then we also are able to create and test our own models, which has a lot of benefits because we’re able to do things a lot more quickly.
For example, we have a client that was working on a project that they had been working on for about five years, and we were able to do the amount of work that they had done in six months with a much, much, much lower cost than they had experienced. So now they’ve just handed all of that over to us because we’re able to do it better, quicker, and cheaper than them. So that was always kind of in our plans to continue to grow and then move more into the full AI lifecycle. But that’s happened a little bit quicker, I think, than we had originally planned, which is awesome.
Raymond Hawkins: Yeah. Growing faster than you expected or finding new lines of business sooner than you expected is pretty cool stuff. All right. So help me, I’m going to ask a question, and if I sound dumb, you can tell me I sound dumb. But I get the data labelling, I get annotating, looking at images, helping an AI engine know what they’re looking at. How far into that part of the process, because I think about this as all a front end or inputs for AI, I know you said you started doing your own data science group, how far into AI modelling do you guys get? Are you really just providing the inputs?
Lauren Bacon Sm…: Yeah, it depends. I mean, most of what we’re doing is providing the client back that fully labelled and QC-ed data, and then they take that and put it into a model that they’ve built. But on the projects where we are doing, actually building the model for them, testing the model, we are doing that full piece for them. And then handing them back that full model with the annotated data in it, and actually a trained model. So it’s not just built tested, but then it’s also the model has been trained.
Raymond Hawkins: And is most of this stuff, you mentioned DOD and IC, is most of what you do tinted, is most of it SCIF related work?
Lauren Bacon Sm…: It’s a mix. So we actually just moved into a brand new office space that we built out, including the secure space. Well, it was in the fall. So we are kind of ramping up on the secured work. And because there’s very limited resources besides us in that area, I think that could quickly take over the unclassed work.
Raymond Hawkins: Gotcha. So let’s tie this back. So you, chief people officer, which by the way love the fact that you were hire number one, employee number two. So as you staff up this work, and I hear it when you talk about, hey, we’re not going to have unique job descriptions for neurodiverse employees versus other employees, we’re going to have a job description and we’re going to mix the teams. I think that’s a super healthy way to look at it. Can you talk to me, as the chief people officer, this is human capital you’re deploying. You’re having to have someone look at and label it, that you’re doing this with human beings. What’s it been like to lead the development of that team and the makeup of that team? Because employee number two, you’ve been in the primary chair from day one. So talk about it from a people perspective.
Lauren Bacon Sm…: Yeah. So we really tried to look at what was actually going to be needed for success in this role and success as a team, and pull back everything else. And a lot of those things that we’re pulling back are barriers for people that have, not just neurodiversity, but people with other disabilities, or just other, I don’t even want to call them limitations, but just diverse backgrounds in the sense of college degrees. We do not require college degrees. We do not require previous IT experience, annotation experience. To be totally honest, we really don’t require any previous working experience at all. I mean, of course there are certain types of experience and certain degrees or certain certifications that definitely will be beneficial and can make a stronger candidate for us, but we’re pretty confident in the fact of that all of the training that we give to our team members is in-house.
So when you do that, and you kind of take somebody that’s more green, and you’re able to really mould them and build them with how we do things and our technical view on things and the knowledge that we want them to have and strategy that we want them to use, and maybe even some of our competitors, we’ve seen that some of our most successful and highest performing team members are people that have no working experience or no relative working experience or degrees. So we’ve really just tried to strip away those things. We’ve stripped away, every job description says strong verbal and written communication skills required. It’s like why? For certain jobs, sure, of course. But for a lot of them, is that actually something that the person needs to be successful?
And in the neurodiversity space, not everyone, but a lot of people, they might struggle making eye contact. They might not be big verbal communicators. So we have really taken away these requirements and expectations and the way that we view and evaluate candidates on how they’re actually going to be successful. And we really just look at what their potential is to be successful in the role versus what their past was. So it’s looking at are they able to follow directions? Are they able to absorb the type of training that we’re going to be doing? Are they able to have that analytical thinking and focus that it takes to do this type of work?
And the way that we evaluate that is that we’ve kind of reversed our recruiting process where the first thing that we do is a technical assessment, versus a lot of people having assessments at the end of a recruiting process. Because for us, like a resume, there’s not really much on a resume that matters or that tells us if they’re going to be a good fit for us or not. That assessment also is two-sided where it lets the candidates see is this the type of work that I could see myself doing every day for 40 hours a week? Because our jobs are full-time, and it’s not for everyone. It can be definitely straining. It can be repetitive. So depending on the type of data, it’s definitely not for everyone. So we want to give them an opportunity to test out the type of work that they would be doing as well.
Raymond Hawkins: Have you seen the series Severance on streaming? I think it’s on Apple+.
Lauren Bacon Sm…: No, but I have heard of it.
Raymond Hawkins: So there’s a group in that building that does data analysis, and you’re talking about all day repetitiveness. It makes me think of that show, or that department in Severance where they’re looking at computer screens eight hours a day, five days a week, just moving data around. It reminds me of what you’re describing. So if you’ve not seen Severance, or our audience hasn’t, it may not resonate, but that’s what I think of when I hear you describing it.
Lauren Bacon Sm…: Now I’m going to have to watch it. Yeah. We’ve really try to focus on ways that our team members can avoid burnout and keeping themselves healthy, with getting up and encouraging them to take at least 10 minutes an hour moving around. We got them all standup desks in our new office space. And we have different decompression rooms that they utilise a lot throughout the day where they can go in, and it’s just a private space where they can sit. And a lot of them have started taking power naps, which we are totally supportive of whatever people need to be productive, which we know is different for everyone, and we really try to enable them to do that.
So we have very flexible schedules. We have some people that come in as early as 5:00, 6:00 AM, and they’re leaving in the afternoon 2:00 PM. We have people come in at 9:30 and stay till 7:00. We also put in place a four day work week for people that want it. So some of our team members work 10 hour days, Monday through Thursday, and they don’t come in on Fridays.
So we tried a few different things with scheduling, and then we said let’s just pull it all back and give them basically as much flexibility as we can. They can come and go whenever they want. They don’t have to tell us when they’re coming and going. And it’s been amazing for them. It’s worked really well. A few of them that were struggling with their schedules before, it just kind of resolved all that. So that’s one of those things that I think it’s like you have to sometimes take a little bit of a risk and try something different too that can make a big impact for your team instead of being so stuck in this is how we’ve always done it, so we can’t even think about doing something different.
Raymond Hawkins: So Lauren, I love thinking about it from the people angle, and I’m going to ask you one more people question. You’ve brought up autism. You used the phrase neurodiverse. I’m going to admit that in my data centre business, we don’t use that term. I don’t know technically probably what it means really well, but I think about people that are ADHD. For me, I’m severely dyslexic. talking on the podcast-
Lauren Bacon Sm…: You’re neurodiverse then.
Raymond Hawkins: … no one would ever know. Yeah, no one would know talking on a podcast. But if you see me try to write something on a whiteboard, and people look at it and they go, “Raymond, I don’t think all those letters are in the right place.” And it’ll take me two or three times. Does it include that? Is the circle that big? That’s what I was trying to spit out, is that-
Lauren Bacon Sm…: Yeah, it is. Yeah. So ADHD, dyslexia, and a few others, dyspraxia, and then there’s another one that I always mispronounce, but they’re different auditory and speech processing disorders. But all of that, there’s a lot more included in neurodiversity than I, to be totally honest, even knew when I first started this role.
So one of our partners that we work with who’s on our advisory board runs a local practise here called Social Grace that helps people that are neurodivergent with different coaching therapy and kind of life skills and job placement. She always talks about we’re all neurodiverse. We all think differently. Our brains all work differently. And that’s really the kind of outlook that we take here is that we just try to create a very inclusive space, and not focus on a disability or neurodiversity. We are just trying to make it as much space as possible for people to be them themselves and be comfortable being themselves and have different ways of doing things, different ways of communicating, different ways of learning, different schedules.
And I think that’s something that everyone should be looking at applying to their companies and their businesses regardless of what your hiring strategy is. Because when you try to treat people as individuals and not just put them into a box and allow the space for them to really be themselves, you are going to get so much more out of them. And you are going to create a place where people want to stay, and you’re not going to have that high turnover where it’s just another place for them to work. If you treat them just another employee and just one of the many, then for them they could get that anywhere.
Raymond Hawkins: Well, I’ll say this, Lauren, I needed you guys in my corner when I was in school because every time I’d get a bad grade, I’d explain to my teacher, “Hey, I just think differently.” And they would generally say, “Raymond, it’s because you didn’t do your homework.” Which there was probably some truth to that too, but it could have been that I thought differently. Could have been a little of both.
So we definitely have a technology audience. I know you and I both said up front, this is not necessarily a tech conversation, but AI is a big technology buzzword. It’s something everybody’s thinking about. We talk about the chips and the computers. And I think it’s in the news today, you see is, I want to make sure I say right, ChatGPT and these engines that can help have conversations and help writing and write poems and write music. AI is certainly, I think, in the news. So tell me what you guys at Enabled Intelligence do. Let’s not get lost too much in widgets and flux capacitors. But what is it that you guys do? How are you helping enable this AI influence on our world today?
Lauren Bacon Sm…: Yeah, you’re right. It is definitely that ChatGPT thing is, every day, I see some new headline about something that it’s doing. It’s definitely a exciting world to be in. And we, I really think, are just that the tip of the iceberg of how this technology is going to be used every day. We’re hearing from different potential clients, even in the commercial space and the medical field, looking at different types of x-rays, looking at insurance companies and paperwork and verifying things. Everyone’s looking at how we can do things quicker and more efficiently and cheaper. And AI is going to enable that in a lot of different fields.
Raymond Hawkins: Can we get the IRS to use AI to get our returns processed faster? Just a suggestion.
Lauren Bacon Sm…: Honestly, probably eventually. I mean-
Raymond Hawkins: You’re in DC. If you know somebody. If you could run by and mention it. I mean, it’s just a suggestion.
Lauren Bacon Sm…: Yeah. I should talk to our CFO about that, and see if maybe you can do some business development and just during tax season, go ahead and reach out.
Raymond Hawkins: Yeah, that’d be good. It’s just a suggestion. I’m just thinking out the box here.
Lauren Bacon Sm…: Yeah. But with the bulk of our business being on the federal national security side of things, that’s part of our mission statement is improving national security and enabling our clients to be able to mobilise these AI tools with the data that we are providing back to them. That often we’re learning projects that have been sitting for years and years because they have not been able to move forward without this first step in the piece. So that’s something that our team members take a lot of pride in as well, is the work that they’re doing. And as much as we can, we share with them about what the client is doing with this and what the mission is so that they’re not just sitting there drawing boxes around images. They know that this work is very impactful in the space we’re in.
Raymond Hawkins: Very cool stuff. Well, Lauren, this has been awesome. I appreciate you chatting with us about it, and helping us understand a little bit about how people are the most important thing, and we all think differently. And how this can help keep our nation secure and keeping our secrets where they need to be. We really, really appreciate you chatting with us. And we’ll remind everybody all this stuff eventually happens in a data centre somewhere, which is why we like talking about it. So Lauren, excited to see the future of Enabled Intelligence. And as you guys grow and serve, not only our country, but your people in your org, pretty cool stuff. And appreciate how y’all think about it. Thank you for joining us today.
Lauren Bacon Sm…: Thanks so much, Raymond.