Choco & OpenAI: Tech For Global Food Distribution Systems

Share this article
Share this article
Prioritise Us on Google
Using OpenAI APIs, Choco processes millions of orders. Credit: Choco
Choco uses AI agents on OpenAI APIs to turn messy orders into structured workflows, cutting manual work and scaling always-on global food distribution

Choco is an AI-powered food distribution platform modernizing how orders move across global supply chains. 

It processes millions of transactions using AI agents built on OpenAI APIs, reducing manual effort across operations. 

The system connects distributors and buyers across major regions while enabling always-on ordering at scale.

Youtube Placeholder
Choco Voice Agent in collaboration with OpenAI

Food distribution for the AI era

Choco operates as a global platform serving more than 21,000 distributors and 100,000 buyers across the US, UK, Europe and the GCC. 

As order volumes scaled, the company faced fragmentation, with requests arriving through emails, SMS, voicemails, images and handwritten notes that had to be manually converted into ERP-ready formats. 

This created operational friction, especially because key ordering logic lived in human memory rather than systems.

“Processing those inputs was the first barrier, but not the hardest one,” says Narbeh Mirzaei, VP Engineering at Choco.

Narbeh Mirzaei, VP Engineering at Choco

“The real problem was implicit context: customer-specific SKU mappings, unit preferences, delivery patterns. 

“That knowledge lived in the heads of order desk reps and we needed to encode it into inference layers that resolve ambiguity at the point of order capture."

Rolling out AI 

To solve this, Choco embedded OpenAI APIs into its core infrastructure, building AI systems that directly process and execute work. 

Its OrderAgent converts multimodal inputs into structured ERP orders, while VoiceAgent enables real-time phone ordering with sub-second latency using the Realtime API. 

These systems are powered by OpenAI models handling text, vision and audio within a unified architecture. 

“The transcription and extraction capabilities gave us a strong foundation,” says Narbeh.

Choco eCommerce combines a modern shopping experience with the power of AI to turn every order into a growth opportunity. Credit: Choco

“The real engineering challenge was building dynamic in-context learning infrastructure, so the system resolves ambiguity against each customer's ordering history and catalog. 

“That's what separates automation from intelligence.”

What’s next for Choco?

Choco now processes more than 8.8 million orders annually, handling more than 200 billion AI tokens in production while reducing manual order entry by up to 50%. 

The platform has delivered double the sales team productivity without added headcount, maintained error rates below 1–5% and enabled fully 24/7 order intake across its global network. 

These improvements allow teams to scale efficiently while shifting away from manual workflows. 

Looking ahead, Choco is expanding its agent-based systems to further automate sales, commerce and supply chain operations across its mid-market global food and beverage customer base.

Company portals

Executives