Eino Provides Corporations With A Decision-Making Platform For Network Upgrades
Major organizations rely on forecasting to plan the necessary budget for infrastructure upgrades. The network improvements are on the order of millions of dollars. Thus, poor planning will result in money being unnecessarily spent. Payman Samadi understood how organizations struggled to do this planning at scale and in a timely manner, creating eino as a result. Eino is a “first-of-its-kind” decision making platform that forecasts demand and optimize resources for capital spending on 5G technologies. The startup is based in New York City.
Frederick Daso: How difficult is it to aggregate such data into a forecasting model, especially in increasing information volumes?
Payman Samadi: We have developed a cloud-based forecasting module that can perform time-series forecasting at scale pretty quickly. One thing to consider is that our tool is mainly used for capital planning; thus, it does not require an instant result and can take time to provide the optimal results.
Daso: What are the limitations of the current methodologies for forecasting in the telecommunications industry?
Samadi: There are two aspects to this problem. First is the forecasting methodologies used for network traffic forecasting. Based on 300+ customer interviews, current operators use a low-resolution shot-gun approach, spreadsheets with generic formulas, or a data scientist developing models. We have shown in our proofs of concept that we can improve network upgrading forecasting accuracy by 35%. The second aspect is that traffic is only one aspect of capital planning. The revenue and cost for each market and even cell site play a part. Thus it's essential to forecast those as well, including external factors such as population growth and data usage growth. Currently, these tasks are happening in silos within different teams that each are optimizing for other goals. Our tool puts them all in one place and enables accurate forecasting + global optimization over the business goal.
Daso: It sounds like current solutions in this space cannot produce actionable insights based on the relationship between demand, population volume, and consumer behavior. What value is unlocked if a new solution could produce said high-fidelity recommendations?
Samadi: The reality is that network operators' revenue has been decreasing in the past ten years; however, the demand for data has increased; thus, they need to continuously add more capacity to the network while making less money. This is a difficult situation that requires more efficiency in capital spending. According to PwC, 25% or $65B of the annual $350B capital spending is wasted. Our tool helps operators to provide the best coverage and speed while having optimal Capex ROI.
Daso: Is the greatest market opportunity is focused on 4G/5G telecommunication upgrades? What could be an adjacent space to move into, provided you're successful in the current one? What would the value of that adjacent market be?
Samadi: Besides the telecom market that is 4G/5G and Wi-Fi planning, we have identified the energy market as a potential new market. After 30+ customer interviews, we realized the similarity between utility and telecom networks. Utility planning is longer-term than telco; thus, accurate forecasts and optimal decision-making are valuable.
Daso: How are the simulations constructed and presented to users when considering all future scenarios? How are these scenarios ranked, and how are recommendations communicated if potential future scenarios conflict with one another?
Samadi: Users build different business cases based on limitations such as budget and network quality. For example, the operators have a $100M 4G budget and a $200M 5G budget for 2021 for specific markets. Also, they want to provide a minimum of 10 Mbps network quality. The tool tells them which cell sites they need to upgrade. Now, the user can change the budget and network quality and see the difference.
Daso: What does eino do to adequately capture, integrate, and leverage spatial analytics to provide better forecasting for your customers? How does eino integrate into companies existing forecasting workflows (i.e., API)? Or is it an external service?
Samadi: Our tool is a decision-making platform that is a full business solution. We integrate with the company's databases and data warehouse and provide our solution either cloud-based or on-premise.
Daso: How are you leveraging your academic background and research experience in approaching the problems you're trying to solve in the industry?
Samadi: I worked as an applied scientist at Columbia University for four years, working with big telecom operators. That was when I came across the problem of network planning efficiency. Then at the Cornell incubator, we made a comprehensive customer discovery and defined the exact problem. Knowing the scientific approach as a scientist, I designed a solution to this problem. We have two accepted patents on our technology.