How AI can help businesses save millions on manual procurement

Procurement, Office Binder on Wooden Desk. On the table colored pencils, pen, notebook paper.[Courtsey]

For decades, manual processing of local purchase orders (LPOs) has been costing businesses across Africa’s supply chains significant amounts of time and resources that could have otherwise been channelled to power growth.

According to the Centre for Advanced Procurement (CAP), companies spend more than 20 hours per week manually entering LPO data from PDFs, emails or blurry handwritten images into Enterprise Resource Planning (ERP) software.

Additionally, depending on the size, CAP observes that companies can spend anything between Sh1,600 and Sh2,000 in processing a single invoice manually.

In contrast, automated processing costs just Sh183 to Sh774.

If a mid-sized company handles just 1,000 invoices monthly, this difference alone could mean more than Sh15 million in immediate, tangible annual savings.

Indeed, not only does manual processing of orders waste time and revenue, but it also results in missed deadlines, piled-up orders, delayed deliveries, a rise in stock-outs and an increase in customer complaints.

Weighing the benefits of automated systems against the disadvantages of manual order management, one may therefore wonder why African businesses have been slow to adopt automated tools.

Traditionally, part of the reason manufacturers and distributors in Africa relied on human effort to key in LPOs is that most of the automation tools that were available in the market only focused on automating tasks such as assembly, packaging, quality control and inventory management.

Today, however, some local software companies are leveraging emerging technologies such as artificial intelligence (AI) to develop tools that can bridge the gap between unstructured documents and structured, actionable orders.

Recently, for example, Solutech launched Eva Docs.ai, an AI engine trained to understand the messy, inconsistent, non-standard LPO formats that dominate African trade.

Leveraging machine learning algorithms, such tools can read LPOs from email or direct upload, map products correctly, extract quantities and prices per customer price lists and tax configurations, as well as post orders straight into ERP systems such as SAP, Odoo and Sage.

As with all disruptive technologies, some may see such solutions for what they are: the engines that will drive productivity, strengthen supply chain resilience and unlock new levels of efficiency.

Other businesses, however, particularly those accustomed to manual workflows, may view such automation tools as a threat to jobs or as a trend that will soon pass.

When all is said and done, in a future where speed and efficiency will be key ingredients for success, one thing will remain clear: procurement automation tools will no longer be a “nice to have” kind of thing.

Instead, they will be the kind of practical innovations that will separate organisations that will achieve progressive growth from those that will not, simply for being unable to move with the speed of change.

For this reason, companies, particularly those that operate in sectors such as the manufacturing and distribution industries that are often characterised by tight margins, complex supply chains and the pressure to deliver orders fast and accurately, must move beyond conversations or hesitation and begin implementing solutions that streamline work.

The writer is the co-founder and chief operations officer, Solutech Ltd