The use of data in today’s digital landscape can be revolutionary for businesses. However, flicking a switch overnight and hoping for the best possible outcome won’t magically turn bytes of data into pots of gold.
In an attempt to get their hands on these proverbial pots of gold, far too many enterprise organisations are caught in a cycle of trying to capture as much data as possible. Before really understanding what business opportunity it is they want to capitalise with the data.
Typically these opportunities are addressed by three main categories. They are:
1) Enhance decision-making: Improve the organisation's ability to make decisions (e.g. which products appeal to customers and which do not?)
2) Optimise processes: make business operations more efficient (e.g. accelerated resolution of customer complaints)
3) Drive additional revenues: redefine their business model and exploit the monetisation of their data assets.
However, having spoken with many enterprise customers both here at Mesh-AI and over the course of my career, it is clear that many are still struggling to maximise the true value of their data assets. As a result, many organisations are embarking on multi-year transformations, uplifting the front end veneers of their digital channels. Without really addressing thorny business opportunities that can be unlocked by harnessing the mountains of contextually rich data that they are sitting upon.
Many of the questions I have heard from customers, just in my first few months at Mesh-AI alone, have tended to align with the following:
- Where do we start with our data transformation?
- How do we make data easily discoverable?
- How do we improve the quality of our data?
- How do we create a self-serve analytics capability?
- What does our future state architecture and operating-model look like?
- Do we need to discard our existing data investments?
- What technical skills do we need in place?
- What do we need to scale ML and AI, in production?
- How can we govern all of these distributed datasets and still move at speed?
Based on these questions and with enterprise organisations striving to maximise business value with their data and the adoption of Artificial Intelligence & Machine Learning capabilities we have decided to create our data driven transformation framework called “Orbital”.
Orbital is Mesh-AI’s transformation framework that enables enterprise organisations to become data driven, by scaling product driven operating models. It involves a defined pathway of steps to unlock valuable business outcomes with cutting edge data, analytics, AI & ML capabilities.
Orbital is our tried-and-tested transformation framework for accelerating, de-risking and amplifying data, analytics, AI & ML capabilities.
It follows a five-phase roadmap of activities and data driven processes for measuring maturity, delivering business valuable capability at scale and ensuring that these efforts are aligned to complex business opportunities for your business.
We believe that the vast majority of enterprise organisations are yet to fully unlock the unbounded potential of their data assets. In order to take their businesses on an accelerated trajectory where AI & ML can deliver unbounded potential and experiences to their customers.
Enterprise organisations often set multi-year strategies that have been incubated for many months, without an evidence based hypothesis validating their assumptions…. Which can often be founded upon sub-optimal data!
We created Orbital to address enterprise procrastination, supercharge transformation and accelerate business value for our customers' exploitation of data.
Orbital has been designed upon Mesh-AI and our people’s real-world experiences of transforming highly regulated, enterprise organisations to become data driven businesses. It also brings together evolving architectural paradigms, such as Data Mesh and automation practices like MLOps & DataOps to ensure that our customers have the people, processes and technology to scale a right sized operating model for their product delivery needs.
Orbital isn’t an all consuming structure of programmatic events. We’ve learnt that transformation is a lot more embryonic and fluid. As such, it has been designed with five key principles in mind:
Lightweight, Yet Comprehensive
It’s a pragmatic set of practices, capabilities and methods that aren’t documentation heavy.
Seed, Then Scale
It aligns to valuable business opportunities & challenges to establish modern data driven solutions before scaled adoption.
Modular and Holistic
It aligns modern people, process & technology approaches to establish a scalable operating model that maximises business value with data, ML & AI.
Battle Hardened
We’ve road-tested these pathways with highly regulated enterprise organisations and built them into Orbital.
Meticulous & Measurable
We baseline performance using subjective and objective data-points, tracking maturity curves and bottlenecks at regular junctures.
Peter Drucker once said: “You can’t improve what you don’t measure”. We have taken that mindset and built it into our Orbital framework. Over the coming months, we will be releasing a maturity model that we use to measure the quality of your data products. This is a self-assessment capability model, that will measure multiple dimensions across your organisation's data lifecycle.
That said, organisations are already swimming in a sea of OKRs, KPI’s and business performance metrics. As a result, it’s easy to measure the wrong data points and as a consequence, make bad decisions for your business. With Orbital we try to keep things simple by taking a pragmatic approach to value measurement:
1) Baseline: We baseline your As-Is performance with our assessment framework and capture objective data points from your systems of record. Agreeing on how to calculate actual performance. This removes ambiguity!
2) Correlate: We track performance at frequent intervals to identify improvements and prohibitors to business and technology excellence.
3) Correct & Optimise: We use evidence and facts to implement treatment strategies and qualify their impact with fast and frequent feedback loops, generated by continuous processing and business intelligence dashboards.
4) Improve: We capture failures, learn from them and start activities to evolve new working practices, technical capabilities and product offerings. This enables us to prove and disprove hypotheses in an accelerated fashion.
Orbital follows a structured pathway that provides a tried and tested journey for enterprise organisations to follow, as part of their data transformations on. It consists of a five-phases of execution. Given the modular nature of Orbital, these phases can be executed in isolation, sequentially or as parallel tracks across multiple business lines in unison to support enterprise scale.
The five phases of Orbital are:
Assess
We educate, inform and coach senior executives about the foundations of real data driven transformation. Building an understanding of your business, its challenges and how data, analytics AI & ML can be used to drive value.
Develop
We benchmark maturity, establish joint hypotheses and establish roadmaps & plans of execution to drive new or uplift existing capability areas across people, process & technology.
Build
We deliver cross-cutting changes that drive alignment across your business needs, change portfolio and compelling events. We call these streams of execution, Lighthouse Projects and they aim to establish:
- Modern data platforms and ways of working.
- Tooling and processes that are highly automated and accelerate change.
- Launch new products and services that harness AI & ML in production.
- Measurable business value that can be analysed in real-time.
Test
We take the learnings from our first series of Lighthouse Projects and feed these into your second and tertiary phases of execution. Further seeding iterations to your operating model across business units to suit the demands of your organisation. Constantly using data to inform our decision making processes and feed next best actions as part of your transformation, we scale your technical foundations by applying a demand driven, product oriented approach.
Additional workstreams may include the commencement and creation of data science academies to re-skill existing people in your business. Or publishing open source projects to demonstrate the credibility of your engineering function and attract new talent into your organisation.
Deploy
With Orbital, we encourage continuous learning and improvement. This is made easier in a data driven environment, where data is democratised, highly available and searchable. Whilst on-demand and dynamic compute resources can be launched to continuously innovate and validate new business requirements, products and services for both internal and external customers.
Lighthouse Projects are an accelerator and catalyst for delivering transformational change across your business. We execute Lighthouse Projects to reimagine your organisation's use of data, analytics, AI & ML. We take this approach to guarantee the biggest outcomes in an accelerated time frame. An ideal Lighthouse Project carries these types of characteristics:
- Compelling event or business opportunity
- Time sensitive, milestone pressures
- Combination of heritage & cloud native systems
- Executive support & sponsorship
- Funded initiatives with a measurable business case
- Open minded teams who want to deliver real business change
This allows us to build momentum as part of your transformation agenda, whilst quickly establishing technical capabilities, processes and ways of working that can be applied across your business and wider operating model.
We look forward to sharing more information on Orbital over the coming weeks and months. If you would like to learn more about how this framework can be used to accelerate your Data transformation, AI or ML adoption plans then please get to discuss how we might be able to accelerate and de-risk your plans.
Interested in seeing our latest blogs as soon as they get released? Sign up for our newsletter using the form below, and also follow us on LinkedIn.
.png)
.png)
