Content development
DevelopedStructuring and drafts for professional content.
Model, architecture and workflow
From insight to commercial action: Agentic workflow
How can AI coordinate steps from input to next action?
The project shows how AI can be used in a structured workflow from input and insight to assessment, prioritisation, suggested action and human quality assurance. The workflow is developed as a concept and practical demonstration — not as an autonomous agent in production. The main question is how AI can coordinate steps from input to next prioritised action, with human-in-the-loop throughout.
The solution is developed as a concept prototype and portfolio demonstration — not presented as a production-ready system.
The challenge the project addresses
Many organisations hold large volumes of insight, CRM data, market information, meeting notes and strategic assessments. The problem is that insight often remains passive and is not translated quickly enough into priorities, sales follow-up or leadership decisions.
The model and workflow
The workflow is built around six connected steps — from receiving input to learning and follow-up:
Input and context
Collection or receipt of input and relevant data sources.
Classification
Interpretation and classification of task type and information.
Assessment and prioritisation
Assessment of relevance and selection of the right workflow.
Suggested action
Draft of next concrete action based on structured insight.
Quality assurance
Human-in-the-loop before anything is executed.
Follow-up
Learning and adjustment based on result and feedback.
What has been developed
The project is developed as a concept and practical demonstration — not as a finished production app:
- Workflow model from insight to prioritised action
- Task understanding and workflow routing
- Structuring of insight and context
- Support for content work and opportunity discovery
- Suggestions for contact strategy and outreach drafts
Core steps and components
The workflow distinguishes between different task types:
Market insight
DevelopedPrioritisation and assessment of market signals.
Target prioritisation
DevelopedAssessment and ranking of relevant companies.
Contact strategy
DevelopedSuggestions for next commercial step.
How the workflow works
Interpret input
The workflow classifies what type of task should be solved.
Retrieve context
Relevant insight is retrieved and structured.
Suggest action
Next prioritised step is suggested as a draft.
Quality assure
Human assessment before execution.
Possible application areas and relevance
Commercial development
Sales follow-up, customer insight, market prioritisation and meeting preparation.
For employers
Shows ability to build workflows connecting strategy, insight and execution — without presenting autonomous AI.
Status, maturity and limitations
Developed concept and demonstrated agentic workflow to structure the path from insight to prioritised action. The solution is not presented as an autonomous agent in production or an implemented client system.
Further development and scalable application
Further development of the workflow
- More data sources
- Clearer agent roles
- Decision rules and quality control
- History and learning loops
Adaptation to functions and industries
- Sales leadership support and follow-up engines
- Market prioritisation and recruitment dialogue
- Consultants, founders and commercial teams
Marius Ottesen
Commercial leader and AI strategist
Develops business-oriented models, workflows and concepts at the intersection of commercial leadership, decision support, knowledge work, technology and practical AI.
Explore the project further
Get in touch if you would like to discuss how agentic workflows can be used in commercial development, sales follow-up or knowledge work.