What is a multipersona DSML platform?

Gartner describes the multipersona DSML platform as:

Democratization, Gartner states, “is achieved by bringing the power of DSML to a wider nontechnical and technical audience while hiding complexity “under the hood” by automation and augmentation throughout all phases in the DSML development and operationalization process.” The report further states, “Multipersona DSML platforms have dual-mode characteristics: first, they offer a low-code/no-code user experience to personas that have little or no background in digital technology or expert data science, but who typically have significant subject matter expertise or business domain knowledge. Second, these platforms provide support to more technical personas (typically expert data scientists or data engineers). Nontechnical personas are provided access through a multimodal user interface that offers at least a visual workflow “drag-and-drop” mode and optionally a higher-level guided “step-by-step” mode”.

The vast majority of AI models still fail to be productionized, and accelerating time to value for AI projects is increasingly a priority for businesses already heavily invested in the technology. 

Gartner reports that multipersona DSML platforms enable acceleration, which allows companies to shorten the time to value for DSML, primarily through “more streamlined deployment, integration and operationalization of models.” 

It is Peak’s view that multipersona DSML platforms address many of the challenges faced by businesses looking to deploy AI. By simplifying, and speeding up, the development and productionization of models, the likelihood of value being realized from the adoption of AI is increased. 

It’s not just about building and deploying models, it’s about delivering outcomes from them.

“Even with the rise of the multipersona DSML platforms, there’s still a tendency for platforms to be heavily weighted towards the needs of data scientists and data engineers,” says Richard Potter, Peak’s CEO. “We believe in delivering tangible business outcomes, and for that, business users need to be brought into the conversation and onto the platform.”

Just helping data scientists to build models is not enough. It creates both a technological and cultural gap between data science teams and business users that can be difficult to overcome. There needs to be an environment where the two groups can collaborate to build effective models, ultimately transforming into critical business applications, without the complexity of stitching together an array of disparate technologies.

Peak solves this problem by offering a suite of configurable AI and Decision Intelligence applications, on top of the DSML functionality, aimed at delivering AI-driven outcomes in commercial decision making. It is in this way that we’re able to support business users, alongside data and analytics personas, to deliver tangible results.

Marshalls are turbocharging digital transformation with Peak

Marshalls is the UK’s leading hard landscaping manufacturer of natural stone and innovative concrete products for the construction, home improvement, and landscape markets. They came to Peak with a clear objective to improve efficiency within their bid process but first needed to unify their data and build the necessary infrastructure to successfully deploy AI. 

Peak collaborated with Marshall’s technical and commercial teams to train and deploy an application that integrated with commercial decision makers’ workflows, enabling line of business users to interact with the model on Peak, speed up the sales cycle and increase the volume of sales. Uniting all teams on one platform means that applications are built with an outcome and end user in mind, and creates a feedback loop between technical and commercial users, so models can be iterated and improved.

When asked about the impact of connecting their teams through our multipersona DSML platform Marshall stated it had “turbocharged” their digital transformation. 

*Gartner, “Market Guide for Multipersona Data Science and Machine Learning Platforms”, Pieter den Hamer, Carlie Idoine, et al., May 2, 2022.

GARTNER is a registered trademark and service mark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved. Gartner does not endorse any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation. Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

Decision making in a changing world

As the world around us continues to change, so do consumer behaviors, trends, and expectations. The continued rise of e-commerce and the explosion of direct-to-consumer channels have been accelerated by the COVID-19 pandemic, and being able to make those split-second decisions to keep your customers happy has never been so important.

However, businesses in sectors such as retail and consumer packaged goods (CPG) lack the end-to-end visibility they need to make the best possible decisions – and making siloed decisions across marketing, demand planning, and the supply chain is hampering effectiveness.

Despite the continued acceleration of digital transformation across industries, the way organizations typically make decisions has yet to materially change and adapt with the times. Yes, data – often static and siloed – may be used to help inform a decision, but too often the onus is still with the human alone to form their own judgement. More often than not, those decisions, formed from an incomplete picture, are based on gut feel and intuition – leading to inefficient and inconsistent outcomes.

Decision Intelligence is here to change that. It enables businesses to make cross-functional, organization-wide decisions in a way that simply hasn’t been possible before. With Decision Intelligence, for the first time, AI can now help us make faster, smarter, more consistent decisions across vast amounts of complex data – on a scale and at a pace that’s beyond human capability alone.

Decision Intelligence according to Gartner

Decision Intelligence isn’t something that’s unique to Peak. Last month (Top Trends in Data and Analytics for 2021: Engineering Decision Intelligence, February 2021), Gartner named Engineering Decision Intelligence as a top trend in data and analytics for 2021.

According to the report, “to deal with unprecedented levels of business complexity and uncertainty, organizations must improve their ability to accelerate accurate and highly contextualized decisions. Data and analytics leaders must explore building capabilities to rapidly compose and recompose transparent decision flows.”

The same report states that “decision intelligence supports improved decisions both individually and collectively, by modeling and optimizing their interactions, measuring business impact, learning and adjusting.”

In November 2020, Gartner also published a report titled ‘The Future of Data and Analytics: Reengineering the Decision, 2025’. It states, “Decision making is becoming more connected, more contextual and more continuous. The current state of decision making is unsustainable. We define reengineering decision making as the fundamental rethinking of business decisions to achieve dramatic improvements in critical, contemporary measures of performance, such as business value, cost, quality, service and speed.”

According to the report, 65% of decisions made are more complex (involving more stakeholders or choices) than they were two years ago.

Where does Peak fit?

The three C’s referenced in Gartner’s ‘Reengineering the Decision’ report – connected, contextual and continuous – resonate strongly with our own vision behind Decision Intelligence as an important new category of software.

Peak’s interpretation of Decision Intelligence is focused on driving connected decision making across an organization, connecting previously siloed data sources across multiple areas of a business. This enables decisions to be made which consider the context of how different events, market conditions and decisions affect each other and impact the entire organization. Plus, it all happens continuously, at scale, and is constantly evolving and improving as the AI gets smarter over time.

The Peak platform can be considered to be a pureplay Decision Intelligence platform – the first of its kind – built to help power smarter commercial decisions. It contains a range of configurable applications that deliver tangible outcomes across the value chain, while  also empowering data teams to build and configure their own decision intelligence applications at pace and scale.

As an example, let’s look at Decision Intelligence’s role in retail decision making. Our platform gives Peak customers – who include the likes of ASOS, PepsiCo, and AO – the ability to predict demand more responsively, looking at marketing campaigns and external data, rather than just focusing on what products have already sold.

They can sell unsold products in a way that maximizes margin, targeting only customers who are likely to buy, in a hyper-personalized way, with the touch of a button. And, from here, supply chains and logistics processes are optimized completely, based on demand that retailers can not just predict, but shape to their advantage. The endgame is an entire organization, operating in harmony, working towards a common goal – winning!

A change point in decision making

Decision Intelligence marks, in our view, a change point in how businesses make decisions. One that has the potential to create a truly data-driven future, in which AI can amplify human judgement and knowledge, bringing new opportunities for businesses to grow and become more efficient.

Gartner’s thought leadership is shining a light on the significance of this new decision making paradigm, and it’s shaping up to be a significantly important new category of technology. As the market continues to grow, we look forward to welcoming other players into the arena while continuing to make Decision Intelligence an integral part of our customers’ future success.

What is AI in business?

In terms of AI’s role in a business setting, there are a number of ways that forward-thinking organizations are utilizing, and benefitting from, this exciting technology. Some of the most common applications of AI in business include the continued rise of customer service chatbots to deal with online questions and queries, as well as the growing usage of image recognition tools and software.

One area that is increasingly gaining traction is businesses using machine learning – a subset of AI – in order to gather more value from their data. Modern businesses, particularly those in consumer-facing industries such as retail, are producing larger and larger volumes of transactional data. This, coupled with the decreasing costs of computational power thanks to the cloud, means that the optimization of data with technology has become a much more viable and valuable option for businesses.

They’re using this ML-powered, enhanced view of their information to improve their processes, increase efficiencies, and to improve the experience of their customers. In order to do this successfully – in a way that doesn’t optimize one aspect of a business to the detriment of another – we believe that businesses need an AI System to act as a layer of intelligence that makes their existing systems smarter.

What is an AI System?

At Peak, we believe that we’re at the dawn of an exciting new era in modern technology; the AI era. Just like every business needs an ERP system or a CRM system, we believe that every business will need an AI system in order to remain competitive in the near future. This, in simple terms, allows businesses to gain a more holistic view of their data, wherever it’s from or whatever shape it’s in, and use this data to power predictive and prescriptive insights.

Peak’s AI System and its three core solutions focus on specific business functions, driving tangible outcomes and ROI tied to wider business objectives. For instance, this could be more optimized advertising spend, increased forecasting accuracy, or reduced logistics costs.

However, as well as being able to optimize and improve specific functions of a business in isolation, a key differentiator of the Peak AI System is its ability to act as a profit-driving intelligence layer across an entire business by connecting up these three core solutions.

This allows you to optimize all elements, functions, and existing business systems. We believe that a beneficial AI system should be able to leverage data from across the entire value chain, rather than from individual data silos like the majority of the out-of-the-box AI tools currently available on the market.

An AI System should be able to seamlessly integrate into all of your existing systems – whether that’s CRM, ERP, or finance systems, for example – whilst getting smarter and smarter over time. This allows you to deploy the technology across the entire business, without needing to rip and replace, and without any extensive integration headaches.

Let’s get under the hood of the Peak AI System and look at some of the key things we considered when building it…

AI infrastructure

By leveraging serverless architecture, the Peak AI System has been specifically designed and built for the handling of large volumes of data at huge scale. It’s an enterprise-grade software-as-a-service (SaaS) platform, which places a key focus on data security. It’s designed to handle datasets and solutions of any size, with inputs and outputs streamed in real time with always-on ML models that get smarter over time as they ingest and learn from more and more data. The more you feed the system, the better the results will be. You can find out more about the infrastructure in this blog by our CEO, Richard Potter.

Management of the full AI workflow

An AI System should offer the unique ability to enable businesses to productionize AI, end-to-end, in a single platform. This means handling the entire workflow, from the ingestion of raw data to a fully-deployed, AI-driven business solution integrated back into your existing systems. This new practice of applying DevOps techniques to build, manage and monitor multiple AI/ML workflows is called MLOps.

Once data has been ingested into the Peak AI System successfully – having been screened through our proprietary GDPR algorithm to anonymize any personally identifiable information (PII) – it’s then unified and transformed using a sophisticated suite of data management tools and techniques. From here, AI solutions are applied to solve specific business challenges, with the predictive and prescriptive outputs then provided via APIs for system integration, or exported directly from the AI System’s user-friendly dashboards.

Seamless integration

A key requirement of any AI system should be the ability to manage and operationalize the AI solutions easily and efficiently. The Peak AI System, therefore, has been designed and built with UI and UX at front of mind, allowing end users to easily access and utilize the AI-driven outputs our solutions provide. Each solution can be controlled via an API, enabling the seamless integration with your existing business systems and without the usual headaches and teething problems. We’ve placed a key focus on building the Peak AI System to ensure that it integrates with – and generates value from – all systems, meaning there’s no need to rip and replace your existing systems when looking to introduce AI and ML into your operation.

TL;DR – here are our five steps to AI System success:

1. Input link: ingest data from your business systems

2. AI Infrastructure: store and unify data in our secure and scalable infrastructure

3. AI Studio: Configure AI solutions to meet your business’ needs, or build custom ones

4. Solutions: Activate individual Peak solutions – either standalone or interconnected

5. Output link: Send outputs directly into business systems using APIs

However, we’ve found that the appetite for innovation in areas such as artificial intelligence (AI) and machine learning (ML) is there. In fact, according to research from Forbes, over 60% of CIOs believe AI and ML are the top critical future technologies. 

Why the big fuss? Well, enterprise AI allows the CIO to move away from the days of siloed data and the proliferation of Shadow IT, and instead empowers them to drive tangible outcomes for their business – with a focus on achieving rapid ROI.

However, there are hurdles to overcome when it comes to becoming AI-powered. Here’s our take on five of the biggest CIO pain points and challenges for 2020 that are faced by senior IT professionals, and some of the main barriers they come up against when looking to introduce AI into the enterprise.

1. No time to innovate

Research from IDG Connect has found that the majority of IT leaders (56%) agreed that their business placed more focus on maintaining operations and ‘keeping the lights on’, as opposed to driving innovative business initiatives and delivering transformational improvements. The benefit of introducing a solution like the Peak Decision Intelligence System, for instance, enables the CIO to drive innovation without the added burden of having yet another system to maintain. Simply put, it’s a quick win.

2. Shadow IT

Shadow IT refers to systems that are being introduced to a business by other functions outside of IT – take martech point solutions, for example. Due to the IT teams’ focus on ‘keeping the lights on’ as discussed earlier, innovation is often instead led by the rest of the business; meaning that IT leaders have less visibility on IT projects and systems being managed outside of their own department.

Of course, the concern here is that this leads to silos of poorly-managed business data that can’t be integrated for value. However, by integrating an AI System that spans the entire business, you can drastically reduce the need for Shadow IT and a proliferation of systems.

3. Impact on existing systems

Those CIOs looking to drive innovation or major change in their business will often get cold feet when considering the impact any new solution will have on their existing systems. Years of digital and cloud transformation projects have led to a high number of CIOs becoming bogged down in complex legacy systems and poor architecture, so they spend a lot of time thinking along the lines of “if we do this, how will it affect these 20 other disparate systems?”

The modern CIO should be looking for a way to introduce AI that integrates seamlessly with their current systems if they’re hoping for a quick time to value and a significant business impact.

4. Messy, siloed data

For many CIOs, the idea of introducing something like AI sounds appealing, but there will be major concerns around the current state of their data – “it’s too messy, we need to get everything in order first” is a common line we hear.

However, this is very rarely the case, and messy data does not mean that you can’t get results from AI. It’s better to start sooner rather than later in order to begin delivering value, and data unification across your systems is the first place your AI journey with Peak would start (and trust us, it’s not as scary as it sounds!)

5. Failed projects

Unfortunately, sometimes a transformation or change project may sound great in theory, but is difficult to achieve in reality. A large number of CIOs, therefore, carry the burden of having led expensive, time-consuming, and late (or ultimately unsuccessful) projects.

This can be an issue for two reasons; the first being that they may be hesitant to look seriously at exciting new tech opportunities, and the second being around obtaining the necessary buy-in they need to secure from above.

What CIOs need is to move away from outdated, risky transformation projects – and instead focus on finding the right partner to work with on an ongoing basis. This partner should be able to offer you the technology, people, and resources you need to ensure you are delivering continuous value from your data with AI and Decision Intelligence.

How can Peak help to address these CIO pain points?

Just like every business requires a CRM or finance system, we believe every business needs an AI System to compete and win in the modern era.

However, the last thing a CIO needs is another costly, time-consuming, rip-and-replace project ladened with the usual integration headaches. Peak exists to take the pain out of the process; our product is a central system of intelligence built to deliver the end-to-end AI workflow and make legacy systems smarter.

Our outcome-focused solutions deliver an ROI in months, not years, and we like to start small and think big with AI and data to solve common CIO pain points. What area of your business do you want to start with on your transformational AI journey?

We had the honor of presenting twice at NRF 2020. On Sunday, our CEO Richard Potter was joined on stage by Tom Summerfield, the Head of Commerce at Peak customer Footasylum, and Tom Litchford, the Worldwide Head of Retail Business Development at Amazon Web Services (AWS.) They discussed the notion of cutting through some of the noise and hype surrounding artificial intelligence (AI) and machine learning (ML) for retail, and instead ensuring that businesses are leveraging the technology to deliver strong outcomes and ROI. We were also able to announce our latest AWS accreditation to the world – we’ve achieved Retail Competency status!

View the slides

Just a few days later, it was the turn of our Head of Retail, Mylo Portas, to take to the stage. Despite being one the final talks on the final day of the conference, Mylo’s fantastic presentation drew the AWS booth’s biggest crowd of the entire event! Mylo delivered a 15-minute talk on the vision of the optimal AI tech stack for retail, and how the technology can be utilized to improve connectivity between siloed and disparate business systems.

View the slides

The rest of the conference provided the team with a great opportunity to hear from other leaders from the world of retail. A common theme that appeared throughout many of the talks we attended was around the ongoing investment from retail businesses looking to get closer to their consumers.

“Retail is recognized as one of the most consumer-centric and innovative industries in the world today,” said outgoing NRF Chairman Christopher Baldwin at his opening Sunday keynote.

“As an industry, we’ve invested billions and billions of dollars over the past decade. This massive investment has started to transform our industry, and it has changed the way consumers shop and how they live. Retailers are taking the lead in personalization and new technologies that will give the consumer even more power.”

According to Baldwin, more than 60% of consumers say that retail tech and innovation has positively impacted their high street shopping experience, whilst this figure stands at a huge 80% for e-commerce shoppers. With 83% of consumers admitting that convenience  – one of the three Cs of the connected retail approach – is more important today than it was just three years ago, it’s no surprise that more and more retailers are looking at ways technologies such as AI can help them offer consumers a more optimal, end-to-end journey and experience.

Attending NRF 2020 has been a fantastic experience, and has certainly provided myself and the rest of the team with some great insights. We feel privileged to have been able to present and network with some of the world’s leading retailers, and meet some truly innovative vendors, whose technology complements that of our own. We look forward to attending once again next year to share more customer success stories and gather further key learnings, but for now, here are some of the other key trends and talking points that caught my eye in NYC:

Other key trends from NRF 2020

? Retailers must be able to successfully harness the huge amounts of data flowing through their businesses, whether that’s transactional or customer data, according to Microsoft CEO Satya Nadella.

? Nadella also explained that retail businesses must build their own ‘tech intensity’ and identity – as opposed to all leveraging the same stack and utilizing the same algorithms – in order to gain an edge in 2020.

? Retail employees should feel empowered in 2020, with roles evolving and knowledge increasing in order to create a better shopping experience for customers (John Furner, CEO at Walmart).

?? Sustainability was, naturally, a hot topic at NRF. One talk that was particularly well-received was from Tom Szaky, CEO of TerraCycle, who outlined the ‘circular supply chains of the future.’

? Retailers are getting more creative in terms of an omnichannel customer experience in 2020, uniting in-store and digital experience, and developing new partnerships (Michelle Gass, Kohl’s CEO).