Platform architecture
Protoolbag's 6 products share one core. This page explains in plain words what that means for you.
One core, 6 products
Think of the operating system on your computer: the core is installed once and the applications run on top of it.
Microservice architecture
Each product runs and is deployed as its own service; background work runs in independent services too.
Each product is its own service
PMS, FixFlow, CRM, PPM, PHR and PEM run as separate services and are deployed independently of each other.
Independent background services
IoT data ingestion, integrations, reporting and scheduled jobs, and AI run in services independent of the product screens.
A product update doesn't stop the others
An update to one product does not stop the others; updates to the shared core are released to every product together, as planned releases.
Scales per service
Capacity grows where the load grows; there is no need to scale the whole platform at once.
Cloud and on-prem installation options
The same product, two ways to run it: the cloud to start right away, your own infrastructure for enterprise needs.
Cloud (SaaS)
Start immediately; we operate the infrastructure, the updates and the monitoring.
- Nothing to install, start right away
- Updates and monitoring are on us
- Add products and plans as you grow
On-prem
The same product images can be configured to run on your own infrastructure through runtime configuration; no custom build is needed.
- The same product images as the cloud, on your infrastructure
- Runtime configuration, no custom build
- Installation and operations planned together with our team
For enterprise needs, on-prem installation is planned per project with our team.
Discuss on-prem installation
Our own interface framework: PTB UI Framework
All our products are built with PTB UI Framework, our company-standard component library, and so is this website. Buttons, tables, forms and notifications behave the same on every screen; once you know one screen, the others feel familiar.
- Designed with keyboard use and screen readers in mind
- Light and dark theme, themeable to brand colours
- Turkish, English, German and Italian interface
One core, many products
Sign-in, users, permissions, notifications, reports and workflows live in the shared core; PMS, FixFlow, CRM, PPM, PHR and PEM are applications running on that core. Adding a product needs no new user list and no new password.
- One sign-in and one permission model
- Each product on its own address, with the same account
- Shared improvements reach every product
Customer data kept apart on every table
Each customer's data is kept apart at database level: row-level security (RLS) is on for every table in the database, so we do not rely on application code alone.
- Row-level security (RLS) on every table
- Multi-factor authentication (MFA): TOTP and e-mail verification
- Tenant data export and reversible deletion
Design without code
You configure your processes yourself with the form, workflow, report and page designers; new record types take their place in the menu. AI can draft a module scaffold from your description; scheduled PDF reports and natural-language queries are ready.
- Form and workflow designer
- Report and page designer
- CSV, Excel and PDF export
Integration layer
We provide scheduled REST connectors, incoming and outgoing webhooks, API keys and an MCP server for AI agents. On the Microsoft 365 side, Outlook mail and calendar sync (CRM), Teams sharing, OneDrive file access and Planner sync (PPM) work today.
- REST connector, webhook and API key
- MCP server for AI agents
- e-Invoice (UBL-TR): the provider connection is set up per customer
- Unified Microsoft 365 hub (all M365 tools in one menu)
Coming soon - Direct MQTT device connection
Coming soon
Responsible AI layer
A strong model handles complex work and a fast model everyday tasks; the primary provider is Anthropic Claude, and if a provider has an outage, backup providers (Gemini, Groq) take over. An AI request goes to the AI provider within the user's permission scope and only with the data the request needs; every action that changes data asks for your approval. On-prem AI is assessed per project.
- Permission-scoped tool calls
- Approval for data-changing actions
- Transparent AI usage credits
60+ automations running 24/7 in the background
SLA checks, sales sequences, scoring, syncs, anomaly scans and scheduled reports run in the background. We monitor errors and automations on our own infrastructure; you can follow platform health on the public status page.
- Scheduled jobs and syncs
- Our own error and automation monitoring
- Public status page
Mobile
The native iOS and Android apps for PMS, CRM, PPM and PHR (coming soon) share one mobile core: the same account and permissions as the web and an offline-first data layer. FixFlow and PEM are responsive in the mobile browser.
- Native apps for PMS, CRM, PPM and PHR
- Same account and permissions as the web, offline work
- FixFlow technicians in the mobile browser
- Release on the App Store and Google Play
Coming soon
Data flow from the field to the report
When the products are used together, data flows within one account; different teams see the same event on their own screens.
- Field deviceController, sensor, meter
- IoT gatewayReadings and alarms
- ProtoolbagData kept apart per customer
- AlarmPMS · priority and escalation
- Work orderFixFlow · SLA tracking
- Energy reportPEM · consumption and carbon
Meter data arrives through data import or a data source connection.
Frequently asked questions
Yes. An on-prem installation is planned per project with our team. The same product images as in the cloud can be configured to run on your infrastructure through runtime configuration.
An update to one product does not stop the others; updates to the shared core are released to every product together, as planned releases. Background data ingestion, integrations, reporting and AI work also run in independent services.
The iOS and Android apps of PMS, CRM, PPM and PHR are in testing and being prepared for the App Store and Google Play; the store date is not fixed yet. Until then, every product works in the mobile browser.
Short glossary
Short explanations of the technical terms on this page.
- RLS (row-level security)
- Every row in the database knows who it belongs to; a user can only see the rows of their own company.
- TOTP
- A one-time code from the authenticator app on your phone that changes every 30 seconds.
- MCP
- An open connection standard that lets AI agents talk to software in a controlled way (Model Context Protocol).
- UBL-TR
- The XML document format used for e-invoices in Turkey.
- Webhook
- A notification one system sends to another automatically when something happens (for example, a new record is created).
- Tenant
- Each customer account (company) on the platform, whose data is kept apart from other customers' data.
- Image
- A product packaged ready to run; the same image runs in the cloud and on your own servers.
- Escalation
- Automatically routing an alarm or request to the next person or team when it is not handled in time.
Let us talk architecture with your team
Our technical team answers your security, integration and data isolation questions together with you.