Lancom joins AWS Anthropic reseller programme for Bedrock

Lancom has joined the AWS Anthropic Authorised Reseller Program for Amazon Bedrock, enabling it to support customers using Anthropic's Claude models through AWS's managed foundation model service.

The designation expands Lancom's role in the AWS ecosystem across New Zealand, Australia and other markets, as businesses assess which large language models best suit different tasks.

It can now help customers design, deploy and govern artificial intelligence systems built on Amazon Bedrock with Anthropic models. This gives customers access to multiple LLMs in one environment, rather than relying on a single model for all workloads.

Amazon Bedrock provides access to foundation models through a single application programming interface, removing the need for customers to manage the underlying infrastructure. It has become a common option for companies testing generative AI services while keeping data, security and compliance controls within existing cloud settings.

Priscila Bernardes, Chief Executive Officer of Lancom, linked the reseller status to the company's broader work with AWS and customer AI projects.

"Authorised reseller programmes give us the raw materials for the creation and delivery of customer solutions," said Priscila Bernardes, Chief Executive Officer, Lancom.

Lancom has been an AWS partner for more than a decade and has expanded further into delivering AI projects for customers. Bernardes said adding Anthropic models broadens the range of tools available for different use cases.

"As an AWS partner for more than a decade and having moved decisively into delivering AI for our customers, the more tools we have in our armoury the better," said Bernardes.

Model choice

Technical Director Gregor Blaj said one of Bedrock's main attractions is access to models through a single API, with infrastructure management handled by AWS. That, he said, makes it easier to build generative AI applications in a managed cloud environment while maintaining oversight of data and compliance requirements.

"It equips us to build and scale generative AI applications within a managed AWS environment while maintaining control over data, security, and compliance, while also using the best AI for specific tasks," said Gregor Blaj, Technical Director, Lancom.

Anthropic's Claude family includes several models designed for different workloads. Blaj said the main points of difference are intelligence, speed, cost and the amount of context each model can handle, making model selection a practical as well as technical decision.

"Broadly, the models differ primarily in intelligence, speed, cost, and context-handling capabilities. For example, Claude Haiku is optimised for speed, while Sonnet is best for balanced performance. The flagship, Opus, costs more and is suitable for advanced reasoning and complex tasks like knowledge work, document analysis, customer engagement, software development and agent based workflows," said Blaj.

From tests to use

The reseller status also allows Lancom to guide customers from early generative AI trials into operational deployments. It is working on strategy, architecture, security and governance for AI projects, and has delivered systems in sectors including healthcare, public services and financial services.

That reflects a wider market shift as organisations move beyond proofs of concept and begin integrating generative AI into everyday operations. In regulated industries in particular, buyers are placing greater weight on data handling, governance and compliance as they decide where and how to deploy models.

Bernardes said the market has moved beyond the initial burst of generative AI experimentation and is now focused on operational use with stronger oversight.

"We're seeing a strong move from the initial experimentation that characterised 2024 and 2025 towards embedding generative AI safely into core operations," said Bernardes.

"Being part of the AWS Anthropic Authorized Reseller Program allows us to help customers access powerful AI models through Amazon Bedrock while maintaining strong controls around security, compliance and responsible use," said Bernardes.