Custom Build or Off-the-Shelf or a mix?
My conversation with an experienced user…
I met a distributor online recently. She explained that she owned a Warehouse Management System, purchased a Proof of Delivery Software as a Service, implemented a licensed-to-use Accounting & Distribution System, deployed a customised online Order Management System and that she is struggling with fragmented data and documents across these systems.
There are perpetual reconciliation and accounting errors. Sales people still fulfil phone and email orders manually. There are frequent backorders and many slow-moving stocks in the warehouse. She believes that a central Inventory System that manages all incoming & outgoing transactions and orders will provide a single source visibility, overcome some of her teething problems, and increase productivity in the Order to Fulfilment process. This Inventory System needs to integrate with the existing systems to create a channel where all information is processed through it.
I asked “why don’t you consider an end-to-end supply chain solution or an Inventory System that is available in the market. They are usually feature rich, proven with built in best practices and lesser development costs & time?”.
She replied, “We are in this business for many years – we have our own best practices aligned to customer experience & expectations, I want to be able to modify – our business is dynamic, growing & rather unique, and I am concerned about hidden costs that can creep up to surprise me”. I understood her point about “One Size does not fit All”. But I could sense hesitation in her voice too. The risk to build is high when you do not have a clear technology strategy to achieve the objective.
I replied to her “in this case, I recommend that you understand more about rapid application development on the Cloud to build your own Inventory System and Application Programming Interface (API) Integration to link all your existing systems together. That done cost-effectively & quickly, you can look at new technologies that will deliver greater values to your business. You are taking your organisation forward instead of continually renewing your existing systems”.
In the Industrial 4.0 technological pillar definition, we should already be looking at:
- Cognitive Computing (in other words, automate works driven by Artificial Intelligence that crunch huge amount of relevant data),
- Cloud Computing (move your application to the cloud so that more devices and people can collaborate and use)
- Advanced Robotics, Internet of Things, Data Analytics
I am excited about this problem statement. I believe many readers will think that I am telling their stories. Through the years of operation and deploying technologies to improve productivity and accuracy, we have laid a host of systems that are silo but effective on their own. But do we need to retire and replace them? I bet not unless they are broken. I see an opportunity to treat these systems as “Services”. We will string them together with new add-ons to deliver a complete solution. We will avoid massive organisation onboardings due to major system changes and user resistance. And I am quite sure it will cost lesser.
From here, I intend to talk about what I will propose to the distributor. I hope it will help her develop a perspective about what is possible in the IT world.
Let us go back to building the new add-ons or the Inventory System. I will introduce a rapid application development technology – the High Productivity Application Platform as a Service (hpaPaaS). It is a no and/or low code development and hosting platform on the cloud. It allows users to drag & drop, configure frontend user components including backend business logics, rules, actions, and workflows on the fly in designated sandboxes (protected staging environment) and publish them (both frontend & backend) to production upon successful validations, simply over a web browser. The ones with low coding capability allow developers to insert codes where heavy customisations are required. There are a few hpaPaaS providers in the market, each with their own focus and applicability. I will recommend selecting a provider with practical experience and knowledge in logistics and supply chain application development.
Leveraging a hpaPaaS……
Using an hpaPaaS approach, we will overcome the below typical custom development cons:
- High initial cost (because you will not need a large IT team)
- Long development time (because there is no or a lot lesser coding)
- Long testing cycle to achieve stability (because the mechanisms, under the hpaPaaS hood, are proven and tested like any commercial off the shelf software)
Now, let us discuss API integration to link all the existing systems. I am assuming the existing systems can export and import data naturally. The intent is for data from each silo to receive and send critical data to the Inventory System by the below defined information flow orchestration.
I will introduce my Integration & Document Hub solution. I built it on the cloud with my team for our customers.
Let me summarize a few key points about this solution:
- It is a secured central repository for all your digital documents (tagged by keywords or values to enable smart store/filing & search/retrieval)
- It has machine learning capability to read digital documents and extract required data
- It has blockchain capability to sign, transfer or share encrypted documents leveraging distributed ledgers.
- It manages and govern many to many secured systems integrations in one box
- It has intuitive user interfaces that support administrator troubleshooting
How it all fits together…..
There are many open-source components to fulfill the above design. It will take some technical ingenuity to piece them together and own it.
Let me touch a little on predictive inventory demand. It should help the distributor a lot. Again, there are many solutions in the market. The key success factor is if you have good (relevant) and clean data to train a model to predict your demand accurately. Amazon Forecast is a fully managed service that uses machine learning to combine time series data with additional variables to deliver highly accurate forecasts. Amazon Forecast requires no machine learning experience to get started. We only need to provide historical data. You can try a few models and analyse the outcome to decide a best fit for your organisation.
In summary, …….
Sintellisys is a technology company that builds Artificial Intelligence (AI) & Web applications on the Cloud for our customers. It was founded in 2018. The founders have a combined 50+ years of technology, consulting, and logistics operations experience, and are uniquely equipped to provide innovative solutions to the industry problems.