Akro AI tackles enterprise data bottleneck

The startup targets document-heavy work in highly regulated sectors.

Akro AI Pte Ltd. is helping regulated industries deploy artificial intelligence (AI) by organising unstructured data into automated workflows, reducing tasks that typically take days to as little as 15 minutes.

“We solved interesting problems… but a lot of time was still spent consolidating documents and pulling information together in an accurate and timely manner," Marcus Quek, co-founder and CEO at Akro,  told Singapore Business Review.

Founded in 2025 by Quek and chief technology officer Brian Tang, the Singapore-based startup builds AI workflows for regulated industries such as defence, finance, and logistics.

After leaving the defence sector, Quek said he realised the same challenge existed across industries where sensitive information limits the use of conventional AI tools.

“At Akro, we came to a clear conclusion: Before you can trust AI to automate anything reliably, you must first solve the data problem and the context layer underneath it,” he said via Zoom.

Clients upload documents, including scanned records and handwritten files, from which Akro extracts information from unstructured data.

Users then define a workflow, letting AI agents learn the process from sample data before clients verify the output prior to deployment.

In finance, the platform automates customer onboarding verification and invoice reconciliation. In logistics, it consolidates data from large volumes of spreadsheets and generates reports for management and clients.

Quek said Akro is also working with defence customers but declined to identify them.

The platform can be deployed without an internet connection, making it suitable for organisations with strict security requirements.

Quek said the system could reduce work that typically takes five to 10 days to about 15 to 30 minutes, saving roughly 90% of the time.

Akro raised $904,000 (US$700,000) in pre-seed funding in June, which Quek said would fund AI and machine learning engineering hires and expand graphics processing unit capacity for in-house model development.

“We want Akro to become that trusted data layer and automation layer for regulated industries,” he said.

Although Akro is focused on regulated sectors, Quek said the company ultimately aims to help organisations build specialised AI systems that could handle sensitive work securely and at scale.

S$1 = US$0.77