5 Reasons for Payroll Processing with project b.

The preparatory payroll processing in tax consulting firms and payroll offices is characterized by media discontinuities, manual data collection, and time-consuming inquiries. Clients provide information in various formats: emails with attachments, scanned time sheets, Excel spreadsheets, or handwritten notes. The result? Clerks spend more time preparing data than on the actual payroll.

project b. solves exactly this problem. The AI-powered platform from Munich was specifically developed for payroll offices and tax consulting firms that want to digitize and scale their processes. Rita, the integrated AI, takes over the preliminary data entry and data validation, while the final control remains with the human.

At a glance: project b. vs. traditional preliminary data entry

1. Ends the data chaos: All formats are automatically structured

The core problem in preparatory payroll processing is not the payroll itself. It is the data that must be collected beforehand.

The reality in many firms:

Clients send payroll-relevant information in every conceivable variation:

• Emails with free text ("Ms. Müller was sick last week")

• PDF scans of sick notes

• Excel lists with overtime

• Photos of handwritten time sheets

Clerks must manually screen, interpret, and transfer this information into the payroll system. This not only costs time but is also error-prone.

What project b. does differently:

Rita, the AI from project b., recognizes and extracts payroll-relevant information from all common formats. The technology is based on OCR (Optical Character Recognition) and Natural Language Processing.

• Sick notes are automatically linked with the period and employee

• Time sheets are digitized and checked for plausibility

• Email contents are structured and assigned to the correct data sets

The result: Clean, verified, uniform data records. No typing, no searching in email inboxes, no error-prone manual data entry.

Practical example:

A tax consulting firm with 180 payroll clients received over 400 emails with payroll data monthly. The manual entry took an average of 2.5 working days. With project b., this effort was reduced to 4 hours, as Rita automatically structured 85% of the data.

2. Reduces email ping-pong and inquiries

Ask payroll clerks what their biggest time waster is. The answer is rarely "the payroll". It is the inquiries.

The problem:

• "Which Müller is meant? You have three."

• "The hours from the 15th are still missing."

• "Is this sickness or unpaid leave?"

• "The bank account details do not match our data."

Every open question means an email to the client. Often, follow-up questions, misunderstandings, and delays ensue. During payroll week, the open points pile up.

How project b. solves this:

project b. automatically structures incoming information and assigns it to the corresponding employees. In doing so, Rita recognizes:

• Missing mandatory information (e.g., start date for new hires)

• Implausible data (e.g., 60 overtime hours for part-time employees)

• Contradictions to existing master data

• Incomplete sick notes or leave requests

Instead of writing dozens of emails, clerks see all open points clearly in a cockpit. Clients can directly complement missing information in the portal.

Practical example:

A payroll office with 12 employees processed an average of 120 inquiry emails per week before project b. After implementation, this number dropped to 25, as clients independently clarified open points in the portal.

3. Works as an intelligent layer over existing payroll systems

Many firms shy away from software changes. The training for a new payroll system takes months, client data must be migrated, interfaces need to be set up again. project b. understands this.

The approach:

project b. does not replace DATEV, Agenda, SPS, or other established payroll systems. Instead, the platform acts as an intelligent preprocessing layer that seamlessly integrates into existing processes.

This is how the integration works:

1. Data entry: Clients send information to project b. (portal, email, upload)

2. Processing: Rita structures, validates, and assigns the data

3. Approval: Clerks review and approve in the cockpit

4. Export: Approved changes flow directly into the payroll system

Even downstream processes like payroll slip distribution or DMS synchronization run through project b. The entire process chain is digitized.

Supported integrations:

• DATEV LODAS and Payroll

• Agenda Payroll

• SPS Payroll

• Other systems on request

Practical example:

A firm with a 15-year DATEV history implemented project b. as a preprocessing layer. The existing DATEV processes remained unchanged. After three weeks of training, the team was productive. The cutover effort? Minimal.

4. AI as co-pilot: Full control remains with the payroll clerk

The skepticism towards AI in payroll processing is understandable. Errors have direct consequences: incorrect salaries, issues with clients, and in the worst case, legal problems. Fully automated black-box systems are therefore not an option.

The project b. approach:

Rita makes suggestions. The final decision always lies with the human.

This means concretely:

• Transparency: Every AI action is traceably documented

• Control: Clerks review and confirm changes before adoption

• Revision security: All steps are logged and audit-ready

• Correction capability: Misinterpretations can be overwritten at any time

This "Human-in-the-Loop" principle meets both professional and legal requirements.

What Rita takes over:

Practical example:

Rita recognized a date incorrectly in a scanned sick note (15.03. instead of 13.03.). The clerk saw the suggestion in the cockpit, corrected the date with two clicks, and approved the change. Without the AI preliminary work, she would have had to manually type the document.

5. A platform for all steps of preparatory payroll processing

The fragmentation of tools is an underestimated problem. Client portal here, email inbox there, DMS elsewhere, payroll system somewhere else. Every system change costs time and carries sources of error.

What project b. unites:

project b. offers a single cockpit for the entire preparatory payroll processing:

• Client and employee portals: Self-service for changes to master data, document uploads, transparency about processing status

• CRM and client management: Overview of all clients, deadlines, and deadlines, communication history

• Data validation: Automatic plausibility checks, matching with master data, highlighting anomalies

• On- and offboarding: Structured capture of new employees, checklists, automated notifications

• Absence management: Central capture of vacation, sickness, parental leave

• Payroll slip distribution: Digital provision, GDPR-compliant delivery

The effect: Standardization as a basis for scaling

When all processes run on a single platform, standardization arises. And only standardized processes can be automated and scaled. Firms working with project b. can grow their client base without proportionately hiring more staff.

Practical example:

A growing payroll office took on 40 new clients with a total of 600 employees within 18 months. Without project b., two additional full-time employees would have been needed. With the platform, half a position sufficed for the additional workload.

Who is project b. suitable for?

project b. is aimed at:

• Tax consulting firms with payroll clients who want to digitize their preliminary data entry

• Payroll offices that want to scale without building linear staff

• Accountants and payroll clerks who want to spend less time on admin tasks

Conclusion: More time for consulting, less for admin

The preparatory payroll processing is a bottleneck in many firms. Not due to a lack of competence, but because of inefficient processes. Searching through emails, typing data, asking questions: These are not value-adding activities.

project b. automates exactly these steps. The AI Rita structures the data chaos, the cockpit consolidates all tasks, the connection to DATEV and Co. completes the circle. Control remains with the human, the work is done by the machine.

For firms that want to grow without proportionately hiring more staff, this is the crucial lever. For clerks who want to leave on time, too.

Sources and further links

• project b.: https://project-b.dev

• DATEV: https://www.datev.de

• Federal Chamber of Tax Advisors: https://www.bstbk.de