“Robots are not going to take our jobs”
Much has been made recently of the potential impact of robotics and automation on the human workforce. Many of the large global accounting houses have published studies citing significant shifts in the makeup of organisations in the coming decade, but while the headlines may focus on job displacement, less attention is paid to the new opportunities the likes of AI creates.
Dhatim is an innovative startup dedicated to optimising the costs of companies through the implementation of artificial intelligence and machine learning. The Gallic pioneers focus on the two largest company spends – procurement and HR – to deliver financial administration tools focused on efficacy and efficiency across all sectors.
Dhatim’s worldwide presence has roots in services provided to the telecoms industry, leading to benchmarking by the likes of Gartners for its unique solution on invoice management. “We’ve received global awards so we know that business leaders understand the positive disruption of what we do,” explains Dhatim CMO David Remaud. “Our customers like Vodafone, SNCF and HSBC tell us this.”
Dhatim’s journey towards pre-eminence in the continually evolving arena of AI has seen a company headquartered in Paris branch out with new offices in South America and Asia. “In the past five years AI has become very efficient in several areas,” notes Remaud. “For instance, taking raw data from an enterprise and sorting and classifying it to find the value. Google has shown AI can recognise a cat from an image and now AI can recognise targeted data on a PDF invoice after being developed to significantly automate basic tasks. For example, with accounting, once the data has been extracted you have to locate each side of the invoice to the right account. This was done manually, but AI is now skilled enough to automate that task.”
Dhatim’s product Conciliator is the leading digital assistant for cost optimisation based on AI, and has proven capable of making savings up to 25% in all spend categories through deep analysis of invoices, reconciled with real usage and best practices via its ecosystem of digitised expertise. It’s a timely solution as Remaud warns that back offices and finance departments, regardless of industry sector, are going to be massively impacted by AI due to the sheer volume of manual work involved. Everything related to indirect procurement will become automated. “As part of a company’s digital transformation, Dhatim Conciliator drastically changes the operational model, enabling end-to-end integration of AI, data science, machine learning and deep learning in a seamless workflow,” he asserts.
But how would Remaud address the fears of those workers who feel their jobs are under threat from the further development of AI? “We have this idea from science fiction, but robots are not going to take our jobs and kill us all,” he laughs. “Education is key to explain what any new tech like AI does. The market has a duty to do this. Now we are seeing a true impact in certain areas of work. Today, we at Dhatim believe AI is enhancing the role of workers by taking away the repetitive tasks and leaving more time to work in ways that add value to a business. For example, in procurement, instead of labouring manually to extract data via complex tools, that classification task is completed, leaving you free to make the important analogies and decisions on top of the data. It’s a question of training the staff in how to use the tools AI can offer in order to enhance their job. It may be that with further automation some roles in accounting become more customer service focused to better improve a business.”
Countries at the heart of the outsourcing movement, like India and Madagascar, could be severely impacted because of their exposure to manual work. Meanwhile, Remaud believes countries in Europe can benefit from AI’s ability to deliver more time to not only create value, but make work both easier and happier. While the workers are whistling, how will AI bring about a step change in efficacy and efficiency for companies that embrace it properly? “This is massive,” says Remaud. “If you’re talking pure efficiency, with Conciliator we have up to a 50% increase in productivity. For example, we are working for a chartered accountancy firm with 900 accounts on its books. So, either the manager decides to fire 50% of its staff or embrace AI to manage twice the number of files than they used to. With regards to efficacy, we rate that in terms of enhanced compliance and the quality of data - robots are much better than humans when it comes to ensuring that analysis of detail. So, we see a 100% increase in quality.”
The improved efficiency is a proven development that Dhatim’s CEO Thomas Bourgeois calls “a real revolution”, and one that is already live. “There are billions of invoices the world over and amazingly most are still on paper,” says Bourgeois. “Even the digital PDFs still require a lot of work to be managed. They must be read, controlled and inputted in the correct system for payment. Ten minutes’ work equates to $12 per invoice. Our robot Conciliator totally automates this workflow. It can read any PDF and verify each line so it’s ready to be put into the accounting system and paid. This process can add value to any business in any industry.”
But how do we begin to make sense of the incoming tsunami of big data? Reflecting on that next step, Remaud believes the challenge for businesses will be to harness AI to tap into under-utilised data to provide actionable insights. “We are not yet there,” he concedes, “as many companies have problems with the quality of their historical and current data because it has been manually entered. AI can ingest, classify and automate the process, so you have an ERP with clean data. The next step can then provide real insights where we can start to develop predictive analytics. However, our customers still have a path to travel. The market has work to do to provide clean data to deliver the next stage of progress in the fields of procurement, finance and HR.”
Dhatim cites its partnership with Allianz as crucial to bringing solutions like Conciliator to market. “We also work with independent financial vendors like SAP, and P2P procurement software vendors are really good partners for us because they take our product on board as a module to automate the classification of invoices for their customers,” reveals Remaud. “And we work with the likes of Capgemini and Deloitte to help re-shape the future of the procurement process.”
Customer-wise, Remaus rates Dahtim’s continuing relationship with Vodafone as one of its greatest achievements to date. Dhatim provides savings by automating tasks and checking V odafone’s lead lines across hundreds of millions of connections each year through the implementation of Conciliator. Vodafone’s Procurement Manager Roger Dashwood has been impressed by its capability to deliver strong analytics. “There are lots of revenue assurance systems out there that check suppliers,” he notes. “But, in terms of margin improvement based on automated cross checking and validation of invoices, Conciliator fits what’s needed. It has the people with the skillsets to help you manage and maximise big data analysis. Agile procurement is a pillar of our business and Conciliator contributes to this goal.”
Global energy group ENGIE also uses Dhatim’s Conciliator as the basis for a brand-new software as a service (SaaS) solution on Vertuoz. “We’ve made Conciliator part of the Vertuoz platform,” reveals its Director General Olivier Gresle. “It enables our customers to analyse huge volumes of invoices against an automated rules engine to verify whether the calculation rules have been correctly applied. Energy bills can now be verified automatically, and in volume, significantly saving the time it takes to identify and secure savings in order to certify the validity of the amount billed.”
Looking to build on these successes, Remaud is keen for Dhatim to grow its geographical footprint, but is mindful of expecting too much too soon from AI, which he feels has reached a plateau. Analysts believe it will take another three years to reach the next significant stage of development. Remaud concurs. “We still have a journey to take in training our machines to understand the supplier’s stack, because AI is not about coding. We are aiming to deliver these solutions in many languages so that’s our future focus; targeting analytics and the way we interact with machines which requires natural language processing.”