About AutomatonsX
AutomatonsX combines technology products and services across multiple business segments. Its work spans enterprise automation, SAP, data and industry-specific workflows, including regulated and non-regulated environments.
Amotion worked with AutomatonsX to bring agentic development practices into its teams and help them build custom agents for different product and customer requirements. The engagement connected changes in how software is developed with practical work on what those teams could build.
One approach, different delivery needs
For a company working across products and services, AI adoption needs room for different use cases. A document-processing workflow, a pharmaceutical QA process and an agent supporting on-premises operations each involve different data, decisions and constraints.
The goal was to establish a shared approach to the agentic development lifecycle (ADLC), while helping teams adapt it to their own work. This included learning to use AI in engineering, organising delivery through Linear and developing agents suited to specific business processes.
Learning through the development lifecycle
Amotion trained the team in ADLC and the practical use of Claude Code. Sessions connected AI-assisted development with planning, implementation, review and QA, keeping engineers responsible for understanding the requirement and assessing the result.
Linear training supported the workflow around that engineering work: organising tasks, making progress visible and connecting the people responsible for delivery. Together, the tool and process training gave teams a common basis for applying AI across the lifecycle.
The work also covered how to customise agent behaviour for different products and use cases. Training and guided builds were accompanied by production work on selected use cases, allowing learning and delivery to inform one another.
Agent development grounded in real use cases
AutomatonsX brought the product knowledge and business context. Amotion provided training and guidance on applying agentic approaches within that context. The work covered several related areas.
- Language, documents and enterprise data. Natural language processing, SQL, document processing and data management across structured, semi-structured and unstructured data, customised for the needs of different products and workflows.
- Pharmaceutical QA and human review. Custom agents and human-in-the-loop systems for pharmaceutical and regulatory QA scenarios, supporting the process while retaining human review and decision-making at the relevant points.
- RPA and SAP workflows. Building and customising agents for robotic process automation and SAP scenarios, connected to the enterprise processes already central to AutomatonsX’s work.
- On-premises and long-running agents. On-premises DevOps agents, alongside long-running and ambient agents for data processing and data management, broadening the team’s work beyond individual coding tasks.
Building capability within the team
The work combined training, guided implementation and support for applying the learning. AutomatonsX’s teams brought their knowledge of the products, customers and operating environments; Amotion helped them connect that knowledge with ADLC practices and custom-agent development.
Engineers and domain specialists remained central to directing the work, reviewing agent output and deciding how an approach should fit the use case.
Measurable delivery outcomes
Following training and transformation work with Amotion, AutomatonsX built and launched products and completed customer deliveries sooner than originally expected. Most of these projects had been accepted and their timelines approved before AutomatonsX adopted agentic coding internally.
| Project | Original timeline | Actual delivery |
|---|---|---|
| NLP-to-SQL SAP chatbot | Expected to take over 24 weeks to build and launch | Built and launched within 10 weeks |
| AI-native email-to-ERP invoice automation | Estimated and committed at 28 weeks | Completed in 12 weeks |
| Ground-sales AI chat assistant for a major B2B food retail client in Southeast Asia | Estimated at 20 weeks | Delivered in 8 weeks |
The SAP chatbot uses natural language processing to translate questions into SQL queries. The invoice automation project was completed 16 weeks ahead of its committed timeline. The ground-sales assistant was delivered 12 weeks ahead of its original estimate, and the engagement expanded into multiple phases.
Reach across the organisation
The engagement supported work across more than 20 repositories and teams involving more than 100 people, spanning multiple products, service scenarios and regulated and non-regulated sectors.
- ADLC, Claude Code and Linear training connected engineering practice with delivery workflows.
- Custom-agent work covered language, documents, enterprise data, QA, RPA, SAP and operational scenarios.
- Training and guided builds sat alongside production work on selected use cases.
- Human review remained part of both software development and the agent workflows covered by the engagement.
A foundation for the next use case
For AutomatonsX, the engagement brought together two complementary capabilities: using AI within the software development lifecycle and building agents for the business processes its products and services support.
AutomatonsX reports faster product development, feature releases and bug fixes following the training and transformation work with Amotion. Teams can apply the same practices to the requirements of each product, customer and operating environment.
The company says its ability to commit to and deliver shorter timelines than competitors is helping it win more projects and extend existing customer engagements into additional phases.