Cargill hunts for scientists to use AI and sharpen trade edge

From shipping patterns to sound of shrimp eating, group eyes data to boost profits

Cargill, the agricultural trading titan, is hiring data scientists to find ways to profit from the scraps of information picked up as food commodities flow through its factories, silos and ports.

The move, revealed in job listings posted on its website, is an attempt to update its envied information edge for the digital era.

The 152-year-old group buys and sells tens of millions of tonnes of crops and meats a year using a network spanning the US corn belt to the Black Sea, but the wider availability of data — from weather patterns to ship movements — has diminished the value of inside knowledge of commodity markets.

Now the company is attempting to better exploit the seven petabytes of information in its proprietary data network. Using information from shipping patterns to the sound of shrimp eating, the company believes data scientists can help it turn a bigger profit.

“Want to work at the forefront of artificial intelligence and agriculture? Cargill has a significant presence across most supply chains in the agricultural space,” a listing for a data science leader at the company said. “With that footprint comes massive amounts of data that can inform us about markets and ways to improve our business practices.”

Among the initiatives is machine learning, a branch of artificial intelligence that sifts through vast data sets to find patterns that can guide decisions. The LinkedIn page of Tyler Deutsch, Cargill global data science leader, said that he oversees more than a dozen employees “building machine learning models for the agriculture, food, and commodities trading industries globally”. “HIRING,” he added.

Justin Kershaw, Cargill chief information officer, envisions using machine learning for tasks including finding the best shipping routes, reading satellite images to assess crops’ vigour and interpreting microphone recordings of shrimp to let farmers know when to add more fish feed, one of Cargill’s products.

“Shrimp make a sound when they eat,” Mr Kershaw told the Financial Times. “In the Cargill data platform, we are collecting acoustical information about shrimp and analysing that.”

Machine learning might also help traders navigate futures markets, where half or more of the volumes now come from automated trading systems.

“We believe that human traders will continue to play a critical part in trading, but will be increasingly complemented by leveraging machine learning to identify market moving events more quickly and on a more consistent basis to enable them to make better trading decisions,” Mr Kershaw said.

Traditionally, commodity traders have made money through access to information, control of critical assets such as ports and superior trading capabilities, according to Boston Consulting Group. The spread of digital technology is “putting pressure on all three of these tools,” a report from the consultants said.

Privately held Cargill’s trading business is not immune. Earnings in its origination and processing division have declined in three of the past four quarters.

Share this graphic Richard Payne, a consultant at Accenture, said that a trading edge in agricultural businesses traditionally came “from being active across the physical supply chain and being the first to get and react to emerging information”.

Artificial intelligence “could look at a lot more data and take the human emotion and error out of the situation. This augments today’s process and changes the role of the trader, but it represents a direct challenge to the commodity trader’s DNA and market worth,” said Mr Payne, a former Cargill executive.

Cargill has an uneven record with big technology investments. In 2015 it took an approximately $170m after-tax charge on a $2bn enterprise resource planning system called Tartan, the company said.

Tech giants such as Google have begun to offer machine learning as a tool to businesses. Cargill has a relationship with Google as a technology platform provider and through work with Descartes Labs, a satellite data company in which Cargill has a minority stake.

As for machine learning, “we really are concentrating on building these capabilities internally”, Mr Kershaw said. Cargill may pursue more partnerships after it masters the skills: “When you’re partnering with people, if you know as much as them, it makes better partnerships.”





1 Comment

  • Howdу! Thіs post ϲouldn’t be written аny better!
    Reading this poѕt reminds me of my previous room mate!
    He always kеpt tɑlking about this. I ԝill forward this post to him.
    Fairly certain he wilⅼ have a good read. Thɑnks for sharing!

Leave a Reply