Email
Password
Remember meForgot password?
    Log in with Twitter

article imageQ&A: Artificial intelligence assists with financial forecasting Special

By Tim Sandle     Sep 11, 2018 in Business
As companies start gearing up for 2019 financial forecasting, artificial intelligence solutions at Genpact have outlined how machine learning can be used to improve financial forecast accuracy.
Forecasting is a key step in the business cycle, enabling companies to monitor finances and plan ahead with strategies to drive projected growth and address potential hurdles relative to investor expectations, geopolitical instability, and overall market fluctuations.
Artificial intelligence can help with this business process; for example, creating the right data pipeline using blend traditional metrics and non-traditional metrics; analyzing financial and non-financial assets, discovering business drivers from the data among others.
To understand the advantages more fully, Digital Journal spoke with Vikram Mahidhar, business leader of artificial intelligence solutions at Genpact.
Digital Journal: What role can artificial intelligence play with financial forecasting?
Vikram Mahidhar: With technologies such as AI and machine learning, a system can be trained using supplied historical data, non-traditional data sets such as weather data, and develop models for performing tasks reviewing future datasets, and providing recommendations to decision makers – including calculating financial forecasts. The more data that AI has to work with, the better its results.
Let’s take a simple example: In mainstream conversational AI applications, voice assistants like Siri and Alexa can answer common questions and execute tasks because they can tap into internet search results and extensive user data from emails, online transactions, location, and multiple apps.
When applied to financial forecasting, AI can leverage data beyond regional market information and bring in nontraditional datasets, such as operational data like shipment or GPS data, which can shed light on expected revenue and sales. This increases precision, efficiency, and overall productivity in the finance function. Practically speaking, AI can all but eliminate a traditional bottoms-up forecast – which is where most manual intervention occurs – learning from the data sets and pushing accurate data down to the required levels of detail.
DJ: How accurate are the predictions from AI?
Mahidhar: AI predictions are very accurate, but, also depends on accurate data input, which can hinder forecasting projections. Machines are only as good as they are trained, so expert data engineering is important from the get go. It’s critical that humans with the domain and process expertise program the AI.
Consider this: a one to two percent discrepancy, in financial forecasting, could cost billions of dollars of revenue for a corporation. At Genpact, we have seen clients use machine learning to improve financial forecast accuracy from 95 to 99.5 percent; which can make a big impact on the bottom line. It’s also important to note that as more data becomes available, machine learning enables models to become smarter over time for continuous improvement – and improving overall accuracy.
DJ: Are there any case studies?
Mahidhar: Using machine learning for revenue forecasting, a global technology company with tens of billions in revenue found greater accuracy by analyzing both financial and non-financial assets. With more data to work with, its finance team was better able to connect the dots in multiple ways for more accurate, strategic results. More importantly, this was effort was led the CFO to identify true drivers of business revenue. Beyond having more data to prepare models, the company can now to better identify the factors that had a significant impact on revenue. For instance, the team discovered traditional datasets, such as volume and price, had far less of an impact on business and revenue than originally expected.
By better understanding what information really matters, the company set up data pipelines that integrate into a machine learning solution that has helped increase their forecast accuracy by almost 150 basis points The speed and scalability of AI allows the company to increase productivity and streamline time and resources. Revenue forecasts that used to take three weeks of effort by a team of 100, are now produced in just two days with just two people.
DJ: Will some companies be resistant to such technologies?
Mahidhar: It’s not just about technology, it’s also about culture, and the change management necessary with people and processes to make the technology implementations a success. Introducing technologies like AI and machine learning can be a radical change for finance teams that are used to traditional methods of financial forecasting. This can lead to resistance among staff, who are understandably reluctant to upend deeply rooted processes and procedures with manually-produced reports they’ve relied on for years. It is important that the CFO serve as champions and communicate the benefits to drive buy-in. CFOs should be at the forefront of adoption and use machine-generated forecasts to make operational decisions, and encourage the culture change needed.
DJ: What can be done to challenge these assumptions?
Mahidhar: Reskilling for AI in general is an issue all functions in a company must face. Genpact research from late 2017 about C-suite and senior executives’ views on AI found 82 percent of respondents plan to implement AI-related technologies in the next three years; yet, only 38 percent say they currently provide employees with reskilling options. The good news is that there’s solid proof that the investment can pay off—if companies adopt the right approach. The same research reveals that organizations that deploy AI strategically enjoy advantages ranging from cost reductions and higher productivity to top-line benefits such as increasing revenue and profits, richer customer experiences, and working-capital optimization.
DJ: Which types of platforms are in development?
Mahidhar: Financial forecasting with AI is an emerging area. The key considerations when selecting a solution is a cloud-based technology platform that is modular to easily integrate multiple digital technologies and your existing systems. A flexible architecture, delivered with a mature application program interface provides the best results, and gives companies the ability to easily scale operations and speed its digital transformation.
In a related article, Digital Journal has met with Scott Rottmann, business leader, global enterprise performance management at Genpact, to discuss how digital technology is disrupting the financial sector in general.
More about Artificial intelligence, Finance, Forecasting
More news from