Case study
From complex spreadsheet pricing to an intelligent, controlled proposal workflow
An AI-powered system that reads a quotation request, finds what is missing, builds the vessel, crew and mission model, and calculates costs and dependencies, with human approval on every material decision.
- Client
- A maritime shipping company
- Industry
- Maritime shipping and vessel operations
- Solution
- AI-powered system for complex price proposals
- Scope
- Proposals involving vessels, crew, training, travel and operational dependencies
Proposal 2381 · Charter, Haifa → PiraeusCost breakdown
- Vessel · 14 days48,300
- Crew · 6 people, 3 roles21,900
- Training · 2 days3,400
- Travel and accommodation6,150
- Port fees and associated payments4,720
Public holiday on 12 May · crew rate adjusted
Total84,470
Missing information
- Crew nationality mixrequested
- Departure dateconfirmed
- Port agent at destinationrequested
The challenge
The shipping company received requests for complex quotations involving vessels and crew. Every proposal required calculations across dozens of variables: departure point and destination, mission duration, vessel type, crew size, crew roles, training days, travel costs, public holidays, availability, port conditions, and other operational dependencies.
The information was managed through large spreadsheets built over time. Only highly experienced staff knew which details needed to be checked, where they were located, and how every change affected the final price.
The challenge was not simply calculating a price. It was ensuring each proposal included every cost component, reflected the terms of the mission, and did not miss a critical dependency that could affect profitability or delivery.
The solution
Vayra designed an AI-powered system that turns proposal preparation into a structured, transparent, and controlled workflow.
The system receives a quotation request in natural language or through a guided form, identifies what information is still missing, and leads the user through the required decisions. It uses a rules-based calculation engine and approved pricing data to assemble a consistent proposal, while retaining human review for every material decision.
Systems involved
- Rules-based pricing engine
- Approved pricing data
- Guided forms
- Human approval
How it works
- 01
A quotation request arrives
The system reads the mission details: origin and destination, dates, duration, service type, vessel requirements, and special conditions.
- 02
Missing information is identified
When an important detail is missing, the system flags the gap and requests the information needed, rather than allowing it to be discovered late in the process.
- 03
The vessel, crew, and mission model is built
The system assembles the crew requirements: number of crew members, roles, availability, training, working days, public holidays, travel, accommodation, and related costs.
- 04
Costs and dependencies are calculated
The pricing engine calculates costs based on approved rules and parameters: vessel, route, duration, crew, training, travel, associated payments, and operational conditions.
- 05
Exceptions and risks are flagged
The system highlights unusual conditions, missing data, exceptional costs, or dependencies that could affect the proposal, so the team can review them before sending it.
- 06
A proposal is prepared for approval
At the end of the workflow, a structured proposal is generated, including the cost breakdown, assumptions, and relevant notes. The authorised person reviews, adjusts where necessary, and approves the proposal for the customer.
What changed?
Instead of relying on complex spreadsheets, institutional memory, and manual calculations, the process became a guided workflow.
Knowledge that was spread across files and people was translated into a system that helps the team ask the right questions, calculate consistently, and understand exactly how each proposal was built.
Before
- Spreadsheets built over years, understood only by experienced staff
- Cost components and dependencies discovered late, or not at all
- No way to see how a proposal was assembled
After
- A guided workflow that asks for whatever is missing
- Consistent calculation from approved rules and pricing data
- A cost breakdown, assumptions and notes in every proposal
Key capabilities
- Reading quotation requests in natural language
- Dynamic forms for collecting missing information
- Rules-based calculation using approved pricing data
- Building vessel, route, and crew models
- Accounting for working days, holidays, training, travel, and associated costs
- Identifying exceptions, risks, and operational dependencies
- Generating detailed proposals ready for approval
- Full documentation of assumptions, decisions, and cost components
The result
Proposal preparation does not simply become faster. It becomes a process that can be reviewed, understood, and scaled.
The team can spend less time searching through spreadsheets and checking calculations, and more time on commercial judgement, professional accuracy, and customer service. New employees can also follow a more structured process without relying entirely on knowledge held by one person.
Case studies
Have a process that looks like this?
Tell us about the work that keeps repeating, the information scattered across systems, or the process that depends on one person’s knowledge. We will help you see what could be built and where to start.
Talk to an AI architect
Share a little about your organisation, the workflow and the challenge. We will get back to you for a focused discovery conversation.