AI development / Web development

Building an AI-powered MVP to empower supply chain risk managers

Industry
Supply Chain Network Risk Technology
Services
AI development services
Years
2020
Team
8 devs, 2 designer, 1 PO Proxy, 1 QA, 1 Scrum master

Alchemai

Alchemai is a US-based startup empowering supply chain risk managers to comprehend, anticipate, predict, and simulate network risk with unprecedented speed and accuracy. Using Artificial Intelligence, they reveal hidden patterns, correlations, and connections within supply chain networks, converting supply chain risk into a competitive advantage.

About the project

When Alchemai approached us, they had an ambitious goal: to change the way supply chains are managed. They noticed that existing supply chain risk management systems typically react to problems that have already occurred and as such, are unable to proactively support robust and efficient supply chains.

The idea for the product was based on the direct insights from the market and the potential clients, and the company has already had a clickable Proof of Concept. The goal was to build a product that would be able to find the patterns, correlations, and connections within supply chain networks, use these inputs to manage and predict the supply chains, and that could be delivered to the potential client.

Our tasks

Interviewing the potential users

In order to better understand the value that the product should bring to its final users (and meet their expectations in the future), we wanted to interview them, learn their pains and objectives.

Building and managing the product backlog

The first thing to do after understanding business goals standing behind the idea and the needs of the users was to translate this idea into a product backlog that would organize the process of development. Then, we focused on managing the backlog by redefining priorities based on user feedback, development team progress, and other factors.

Building the backend of the application

The client already had the designers and the frontend team. Our job was to take care of what’s inside the application – from planning its architecture to making sure all the mechanics beneath will meet stringent requirements.

Building the AI module

Building a product that would be able to analyze enormous volumes of data to find the patterns, correlations, and connections within supply chain networks required putting Artificial Intelligence at its heart.

01

Challenges

1. Many factors possibly impacting supply chains

The supply chains can be possibly impacted by numerous factors: geopolitics, delays made on the sub-contractors level, logistics, natural disasters. And this impact can be measured with the value of lost contracts, contractual penalties, or storage costs – the pain is real. The challenge for the supply chain managers is to predict possible delays and react accordingly.

2. Inefficient supply chain risk management systems

Existing supply chain risk management systems often simply react to the problems that have already occurred and as such, they can’t support supply chain risk managers efficiently.

3. Meeting users expectations and requirements

From the very beginning, the product built by Alchemai was meant to solve a specific problem. Insights from the market and its target customers were incorporated in the process of design & development from the very beginning. The challenge was to translate these insights into requirements organized in a solid backlog that would help to manage the whole process.

Neoteric is definitely one of the best software development contractors that I’ve worked with over the past 20 years. Their responsiveness to deadlines is integral in the success of our project, and they never fail to deliver. They take our business seriously, and we’re always reminded of that through their transparency and accountability.

Jay Modi, CTO at Alchemai
Source: Clutch.co

02

Solutions

1. Interviewing the potential users

Before we built the backlog, we wanted to talk to potential users of the product, the supply chain managers, and find out more about their problems and needs. We did it in cooperation with the client’s team. Thanks to that, we were able to prioritize the features and start building the MVP with the ones that were the most crucial from the users’ perspective.

2. PO Proxy joining the team to support Product Owner

In order to support the Product Owner with managing the product and its development, the PO Proxy has joined the team. The role of a PO Proxy was to help the Product Owner manage the backlog, set priorities, and define acceptance criteria.

3. Using AI to predict supply chain risks and their impact

In order to manage the supply processes efficiently, it’s important to know whether there will be any delays and if they occur – to be aware of where in the supply chain it will happen and how it impacts the customer’s revenue. Utilizing Machine Learning, our models are able to predict different risks related to supply chains and their effects.

4. Cooperating closely with the client’s internal team

As the client already had the designers and a frontend team, it was important to team up with them and cooperate closely. From the beginning, we established a true partnership, working as one team. We went through the research phase together, shared the conclusions, and planned the general rules of cooperation. We were working in Agile sprints that we planned together with all the team members. Despite having different areas of expertise, members from both teams did not hesitate to suggest improvements to others, both in terms of technological and organizational issues. Moreover, we were updating ourselves during daily standups, and documenting the results in cyclical periods.
Apart from that, we actively participated in discussions with other members, supporting in making product decisions and cooperating on solutions for the AI module with the client’s stakeholders.

Technology we used

AI Predictive Models

Java

Python

03

Project Results

Supporting a sales team with accurate order value estimates

The development of the PoC took two weeks and it helped validate the initial hypothesis about the factors influencing price fluctuation on the market. The result PoC delivered assured the client that AI can be useful in solving his business case and that it’s worth continuing the development. With the PoC the client received a model providing estimates of the order value and delivering the estimates to the sales team. The process of getting an estimate includes exporting the order as a CSV file, sending the file to a specified email address – the email then triggers a robot that retrieves the missing data from the database, starts the predictive model, and gets the estimate. The estimate is then sent as an email to the sales team.

Decreasing margin of error

The solution will support the client’s sales team in their everyday work, making it easier to provide calculations for the customers. The end customers will get more accurate estimates, and they won’t have to plan for a big margin of error between the due amount estimated on the day of placing the order and the actual one that they have to pay upon delivery.

Conclusion

The project for Alchemai is interesting and challenging. Interesting because we are dealing with a matter which is not yet well developed on the market and we come to many conclusions and ideas on our own. We cannot simply look at how others do it and do it better. We come up with the trend ourselves. On the other hand, it is an interesting challenge from an organizational point of view. People from several countries, different time zones, and different cultures work on the project. Combining all these aspects is exciting and stimulating.

Maciej Borkowski, Product Owner Proxy at Neoteric

Do you like our work?

Let's talk about your project!

Discover our
other projects