Three Trading Platforms on Model-driven and Portal Power Apps
Global agricultural merchant
In short
Can Power Apps run high-volume trading platforms for internal and external users?
A trading desk loses time and control when its tools are slow, scattered or closed to the brokers it deals with. The client needed platforms that traders inside the company and brokers outside it could both use at the pace of daily business. We developed three, using model-driven and portal Power Apps.
Results in numbers
- trading platforms
- 3
- traders and brokers
- 100+
- transactions a day
- 500+
At a glance
- 3 large-scale trading platforms
- 100+ internal traders and external brokers
- 500+ transactions a day per app
The client
The client is a global merchant and processor of agricultural goods. The work sat in its Global Process Automation Team, which oversees automation across regions. The platforms served more than 100 internal traders and external brokers.
What it was costing them
- Transaction load: each platform handled more than 500 transactions a day, so manual handling does not scale.
- Two audiences: staff inside the company and brokers outside it need different views of related data.
- Control risk: trading data needs role-based access, and a mistake in who sees what is a business risk.
- Change speed: requirements moved as business teams refined them, so the backlog needed tight management.
What we built
We built three platforms, with model-driven apps for internal users and portals for external brokers, refined with the business through agile sprints.
- One platform per business need: three large-scale solutions instead of one stretched app.
- Internal and external users covered: model-driven apps serve traders, and portal apps give brokers their own entry point.
- Built for volume: each app handled more than 500 transactions a day.
- Requirements that stay current: backlog refinement in Azure DevOps kept sprints tied to what the business needed.
- Stakeholders in the loop: requirements gathering, solution architecture, regular updates and agile testing validated each release.
- Part of a wider programme: the platforms sat alongside the team's audits of automation projects in other regions.
Results
- Three large-scale trading platforms developed during 2022.
- Used by more than 100 internal traders and external brokers.
- Each app facilitated more than 500 transactions a day.
The risk most teams miss
The month three risk is scale and access. Portal users are external, so permissions, table security and licences must be right from the start, and daily transaction volume turns small design shortcuts into slow screens and large tables. We separated the internal and external experiences and managed the backlog in sprints. Today we also plan data retention and capacity monitoring from day one.
What we would add today
Improvements we would add on top of what is live.
- Use Dataverse row-level security by business unit and field-level security on prices and counterparties, with Power Pages table permissions for brokers.
- Sign brokers in with Microsoft Entra External ID instead of local portal accounts.
- Archive old transactions with Dataverse long-term retention: at 500 transactions a day, one app adds well over 100,000 rows a year.
- Release through Power Platform pipelines into Managed Environments grouped in an environment group, with connection references on service accounts.
- Add a Copilot Studio agent for traders to ask for the status of a deal, honouring Dataverse security.
- Feed desk reporting to Power BI through Fabric, so volumes and cycle times are visible by platform.
Technology
Power Apps (model-driven), Power Apps portals, Dataverse, Azure DevOps.
Tell us how many people trade, approve or broker through spreadsheets and email today, and we will tell you which platform they need.
Products and capabilities
- Power Apps
- Power Pages
- Dataverse
- Azure DevOps
- model-driven app
- trading platform
- external portal
The service behind this work
Power Platform developmentThe service behind this work
Dynamics 365More on this industry
Manufacturing and distribution
Related use cases
Sales and customer serviceCustomer Portal for Order Placement and Tracking
When customers order by phone or email, the sales office types each order in again, answers every call asking where a delivery is, and finds mistakes after the lorry has left. We built a customer portal where approved customers place and track orders without the sales office in the middle.
CRM with a Standard Sales Process, Customer 360 and Quotes
Customer data at the client lived in personal notes and spreadsheets, prospects were validated by hand, and every salesperson worked slightly differently. Management could not make decisions from data that did not exist in one place. We built a CRM with one sales process, a Customer 360 view and a semi-automated quotation process.
Mini CRM Options on SharePoint or Dataverse for Teams
A client list kept in spreadsheets has no history, no clear owner for each relationship and no shared view of who knows whom. A full CRM is too much for a team that needs a simple one. We compared three light options for a mini CRM, so the team could choose with the trade-offs in front of them.
- 13 hrs
- model-driven build
- 42 hrs
- per canvas build
Have a Microsoft project in mind?
Tell us what you need built, fixed or supported. A senior consultant replies within one business day.