Camera operators filming in a brightly lit TV studio

News

Company NewsArticle

Complexica Launches Complexica Automations, its Agentic AI Business Unit

30 October 2026

ADELAIDE, AUSTRALIA, 30 October 2026 - Complexica Pty Ltd, a leading provider of Artificial Intelligence software for optimising sales, marketing and supply chain decisions, announced today the launch of Complexica Automations, a new business unit that designs, builds and operates AI agents that carry out multi-step business processes inside customers' systems, with their people approving the steps that matter.

Chat assistants answer questions, and rule-based tools such as robotic process automation (RPA) repeat the same fixed steps until the input or the process changes. An agent is given a goal instead: it plans the work, acts in a company's systems, checks the result and keeps going until the job is done, within limits the company sets. Complexica Automations builds the system around the model that makes this safe to run in an enterprise, including connections to systems such as the ERP, CRM, inbox and shared files, memory, testing, rollback, the ability to stop, permissions, human approvals and a full audit trail.

Complexica as its own first customer

Before offering agents to customers, Complexica put them to work on its own processes. Ten agents have been built inside the company, seven of them already in use and three in development, and every agent the unit offers has first run on Complexica's own work.

Ten agents, built and proven inside Complexica first: Support, Dev, QA, CPX Website, Commercial, Legal and Demo Design Agents, plus Customer Support, Cyber and Legacy Management Agents in development

The agents already working inside Complexica include:

  • Support, from ticket to fix: the Support Agent investigates a ticket in a sandboxed, read-only environment and works the diagnosis through with an engineer in Slack. If a code change is needed, the Dev Agent makes the approved fix and the QA Agent checks it before anything goes live. An engineer approves before anything changes.
  • Legal questions and contract reviews: the Legal Agent answers questions from signed agreements, redlines and emails, and reviews draft contracts against Complexica's standard positions to return a memo and a tracked-changes redline. Every conclusion links back to its source, and high-risk and commercial decisions stay with the authorised reviewer.
  • Website changes from one Slack message: the CPX Website Agent rebuilt complexica.com from a hosted CMS into native code, testing every page, and now handles every change to the site, checking each affected page before and after it goes live. This announcement was published to complexica.com by the CPX Website Agent from a single request in Slack.
  • Commercial proposals and demos: the Commercial Agent drafts and reviews scope proposals, and the Demo Design Agent makes and verifies changes to demos.

Three further agents are in development: a Customer Support Agent that diagnoses and resolves customer issues, a Cyber Agent that finds where a business is exposed and proves which security vulnerabilities are real before an engineer approves each fix, and a Legacy Management Agent that updates and maintains legacy codebases, already used to turn a car-distribution optimisation system from a desktop application into a modern web app.

"Every agent we offer has run on our own work first," said Spiros Kaye, Head of AI Automation at Complexica. "Using them on our own support, legal and website work showed us where an agent can be trusted to act, where a person needs to approve, and what the system around the model has to do to make that safe: connect to the right systems, remember what it has learned, test its own work, roll back and stop. That system is what turns a capable model into something a business can actually run, and it is the layer we build for our customers."

The same patterns, in customers' businesses

Each internal agent follows a pattern that applies well beyond Complexica. The support pattern fits customer self-service and any request that has to be investigated across several systems before someone can reply. The legal pattern fits supplier document checks, invoice matching and approval drafting. The website pattern fits living standard operating procedures, reporting and commentary, and the legacy pattern fits risky spreadsheets rebuilt as governed apps with approvals and audit trails. Complexica Automations is also in discovery with customers on knowledge and answers, operations monitoring, market and disruption watch, process mining and contractor compliance. Complexica's Industries page now shows where agents fit in each of the sectors it serves, from CPG manufacturing and retail to ports and logistics, pharmaceuticals, wholesale and distribution, and financial services.

"Most organisations have moved past asking whether AI is useful. The question now is which processes it can take on safely, and what that is worth," said Adam Michalewicz, Manager, AI Automation at Complexica. "We start with one process, look at the systems and approvals involved, and tell the customer honestly whether an agent is the right tool. When it is, we design, build and run it with them, so they get a working process with measurable results rather than another experiment."

One team, from first conversation to a running agent

Complexica Automations works with customers in five steps: discovering the processes where an agent can take on multi-step work, designing which systems it reads and acts in and where a person approves, building and validating the agent with every release versioned and evaluated before it goes live, running it inside the customer's IT with permissions, approvals for high-risk steps and traces and alerts on every run, and reviewing finished work on a schedule so each job makes the next one easier. Every deployment is governed by Complexica's ISO 27001:2022 certified information security management system, and the unit is model-agnostic, working with any frontier model so customers are not tied to one AI provider. The agents build on more than 25 years of applied AI research at Complexica.

Spiros Kaye leads the technical side of the business unit, taking frontier AI research through to working systems, and Adam Michalewicz leads the commercial side, helping customers identify where automation can deliver the greatest value and scoping those engagements.

To learn more, visit the Complexica Automations page or book a discovery call to talk through one process with the team that builds the agents.

Contact

To schedule a demo or learn more about our software products, please contact us:

Request a Demo

Some of our Customers