Projects

1The problem

Fenyx was running more and more projects through partner companies, with no central place to manage them. Project details sat in phone calls, emails and people’s heads.

Partners handed in work proofs on paper. Staff typed them in and compared them with invoices, financial plans and the work on site by hand. The finance systems were rigid and not connected to project work.

2What I built

• A central platform for project details, responsibilities, leads and partner information.
• Digital work-proof submission for partners, replacing paper.
• Checks that compare work proofs with invoices, the financial plan and what happened on site.
• Finance workflows connected to project management, with budget tracking.

1The problem

Candidate and customer data sat in the CRM and in the recruiting platform, and people moved it between the two by hand.

Records were fragmented, which slowed down the teams that relied on them.

2What I built

• Automations that connect the CRM with the internal platform.
• A clean-up and enrichment of candidate and customer data.
• Workflows that hand work from one department to the next.
• Database analysis and maintenance to keep the data clean.
• Invoicing automation with n8n, Pipedrive, Qonto, Google Sheets and Slack.

1The problem

Acker’s backend ran on an older framework and had built up technical debt.

The team was migrating it to LoopBack 4 and, in the same effort, moving from a monolith towards separate services.

2What I did

• Worked on the migration from LoopBack 3 to LoopBack 4.
• Helped move the monolith towards microservices and domain-driven design.
• Planned the work around technical dependencies, refactored code and reduced technical debt.
• Took part in code reviews, effort estimates and troubleshooting.

1The problem

LoadUp held around 60k applications from drivers. An application records one moment: one person applying for one job. It does not describe the person in a form that software can compare with open jobs.

2What I built

• A process that turns applications into applicant profiles with structured fields.
• Profiles prepared so the matching algorithms can compare them with open jobs.

1The problem

Some drivers would rather send a message than take a call or fill in a web form.

The logistics companies, often without HR staff, could not answer every applicant by hand at any hour.

2What I built

• A WhatsApp contact for LoadUp that drivers can write to at any time.
• Drivers can ask questions, update their profile and say what kind of job they want.
• The service passes those requests to the matching engine and sends back suitable jobs.
• Drivers complete the application inside the chat, without switching to a form.

1The problem

LoadUp recruits truck drivers for logistics companies, many of them mid-sized firms without an HR department.

Drivers are on the road, and a long online application form was a barrier. Applicants were dropping out of the existing online funnel.

2What I built

• Paula, a phone agent that takes drivers through the qualification questions. I built it from the first concept to production.
• It runs on a voice platform we chose over Vapi because it meets GDPR requirements.
• Paula takes inbound calls from drivers and makes outbound calls. During the call she reads and updates the driver’s profile and queries the database for suitable jobs.
• The driver can apply on the call, and the application is delivered to the client company. It is the same logic as the WhatsApp service, in voice.
• Campaigns ran in six countries.