Meet Algo, the AI enabled virtual business analyst. Algomus has created the bot to help grow the collaborative intelligence of enterprise with people and AI agents working hand in hand to deliver exponential business outcomes. Algo is the world’s first AI enabled conversational Business Analyst and an integrated workflow environment of its kind. Algo is always on. It mines Big Data and develops an internal representation and higher-level understanding to help business people get answers to their questions quickly, easily and accurately. Algo learns to perform tedious analytic tasks to give business analysts more time to be strategic and creative.
At the AI in Industrial Automation Summit this June 26 - 27, the team will be joining RE•WORK to exhibit Algo and chat to attendees about their current work. In advance of the summit, I spoke with Nikki Hallgrimsdottir, Co-Founder and Chief Evangelist to hear about her current work and their use of AI and deep learning.
Give me an overview of your work at Algomus
I have had the great pleasure for the past year and a half to be working with a group of amazing people and some awesome customers to develop a conversational assistant - Algo - that takes on the role of a business analyst at our client companies. Algo is both a robot that can be taught to perform tasks such as making calculations, forecasting, creating orders, doing research, fetching and sending out reports, etc. and a platform for his colleagues to collaborate using their company's data as a driver for increased profit and cost savings. We primarily work with Industrial Manufacturers, CPG Companies, Distributors and Retailers since that is where a lot of our team's expertise hails from in areas such as Inventory Management, Assortment and Pricing Optimization and other predictive and prescriptive analytics applications. We believe when humans and machines work together they can deliver superior results; we like to call it Collaborative Intelligence. In my current role I am definitely fulfilling a dream working very closely with our developers and our customers to utilize what we are building to transform their business, all the while getting to research and attend industry-leading conferences to make sure we are offering state of the art solutions and sharing what we are doing with the world.
How did you begin your work in AI and more specifically in AI assistants?
I have been working in the area of Industrial Automation for a few years, and then spent a few years in the software business - but still focused on hardware technology. I worked with technology companies such as Apple, Google, Intel, etc. and being in Silicon Valley I was acutely aware of the advances being made in AI and how software companies doing "AI for x" were popping up in SF and elsewhere.
What are the main challenges in creating your AI-powered virtual business analyst? How are you working to overcome these?
One of the major challenges that all AI companies face is the availability of good training data and the scalability of AI applications built on data. Having to train each implementation on customer data sets that are never ideal doesn't scale well and is costly, however, it is a definite prerequisite to creating effective AI programs that work in real life situations. We have been lucky to work with some great customers early on in our development so we have been able to build out models and applications that can then be generalized to future customers, but it takes time and real-world data is always tricky especially in the early stages when companies are not collecting data with AI model training in mind. Thankfully our approach of building training and feedback into the use of the tool allows it to become more sophisticated over time. As we grow and more users work with Algo the data challenge gets smaller every day.
How have recent advances in AI helped your research?
The availability of open source packages in ML/AI has been great for the field. We like to stay vendor agnostic when it comes to our tech stack, not overly relying on a single provider or platform to stay flexible and innovative. Things move fast and we believe it is important to focus on the results and to do that you need an ever-expanding toolbox to work with.
How are you using AI for a positive impact?
We have found that when you give people tools that are a pleasure to interact with and make people's jobs easier, their job satisfaction goes up. Our users tell us time and time again that the time they save with Algo doing their busywork and making it easy for them to use their data to be more analytical and strategic has made their jobs more enjoyable. We believe the future of work isn't one where robots and AI take over people's jobs, rather redefining the tasks that make up our workday. Using AI to turn rote tasks and busywork into automated workflows letting people serve the human functions that are more creative and enjoyable is a positive for the workforce in the long run.
What developments of AI are you most excited for, and which other industries do you think will be most impacted?
I'm really excited about advances in NLP and NLG technology enabling better conversational interfaces and AR/VR as a way to interact with technology. The point and click way that we use software today will mostly be replaced by type, voice, touch, or thought based interfaces where no one has to waste time opening windows, clicking through menus, dragging and dropping things etc. instead we can just ask for what we are looking for or want to do and interact with technology in a much more fluid way.
AI and machine learning raise many ethical concerns such as bias, security & privacy amongst others. What are your opinions on this and what can be done to avoid biased machines?
It's all about the data that AI machines are trained on, and it is very important not to take for granted that data can be biased (not the data itself, but the way it is collected and used). When dealing with any kind of personal information, or using AI to make decision that impact people's lives it is that much more important to make sure that we are aware of and correct for imbalances that will cause bias in training data and try to correct for it. It is out responsibility as developers of technology to think about how it will affect our society in the future and what type of society we want to live in. One way to combat bias in AI is to diversify the voices that are involved in creating it and making sure test datasets are weighted for as much bias removal as possible. Getting underrepresented people to participate in the conversation is also really important, which is why I am personally involved in groups like Humans for AI and Accel.ai which aim to demystify the field and open it up to women and people of color which are typically affected most by institutional bias.
As a woman working in the field, have you faced any biases, challenges or obstacles?
Honestly, no not really. I have been lucky to work with great colleagues and customers in this space, and haven't experienced any issues or setbacks due to my gender as far as I can tell. Back in the day as a Field Sales Engineer, I was acutely aware that I was a woman in a man's world most of the time, so this has been a refreshing change. I understand there is still a lot of inequality in the AI field, particularly I think on the Academic/Research side, but on the business side of things I have to say I have been pleasantly surprised.
What can we do to encourage more women into AI?
I think we just have to continue to break down gender stereotypes and gender-specific roles in general, especially in early education, which will ultimately lead to more equality in STEM and other fields in the future. The fact that one can find ABC's for Boys with images of "all of boys' favorite things--airplanes [and] dump trucks" and ABC's for Girls with "butterflies, castles, and more!" books on Amazon baffles me.
We start at such an early age implicitly telling boys and girls what occupations to gravitate towards. Currently, AI is a pretty advanced field that requires a pre-requisite technology background that is currently underrepresented by both women and minorities, so it is impossible to really level the playing field today. However, I think in the future this field will become much more mainstream requiring less advanced education and therefore a shorter path to entry, and in parallel, we are steadily increasing the number of women graduating in fields that can lead to working in AI. In the short run, I would love to see more of the women in this field publicising their work and speaking at conferences etc, to make it more approachable for other women to enter the field and take this path.