Services Services How It Works Voice AI Sample Audit Internal Tools Resources Scenarios Proof of Work FAQ Blog Company Sectoren About Contact Pricing Client Login Book a Strategy Call
Proof of Work

What we have built and what happened.

We do not publish client names without permission. These are anonymised summaries of real engagements — structured to show the problem, what was built, and what was measured.

Four representative engagements.

Real Estate 15–30 employees Completed

Lead qualification and follow-up automation

The problem

A regional real estate agency was receiving 80 to 120 inbound inquiries per week. Consultants responded manually to all of them regardless of intent or budget fit. High-value leads were treated the same as casual browsers. Response time averaged 4.2 days — well above what competitors were delivering.

What was built

A lead qualification system that scored incoming inquiries against firmographic and behavioural signals, and triggered differentiated follow-up sequences based on score. High-intent leads received immediate personalised outreach. Low-intent leads entered a nurture flow requiring no manual effort.

Integrations

CRM (Pipedrive), email platform, Funda inquiry feed, calendar booking system.

87%
qualification accuracy at 60 days
4.2d → 1d
response time to high-intent leads
11 hrs
saved per consultant per week
Financial Services 30–60 employees Completed

Internal knowledge retrieval system

The problem

A financial advisory firm had regulatory documentation, product guides, and client policy terms distributed across SharePoint, email threads, and individual folders. Advisors were spending 30 to 45 minutes per query finding accurate answers to client questions — often escalating to a senior colleague who had memorised the relevant policy.

What was built

An internal knowledge agent connected to their document sources. Advisors query in natural language and receive accurate, sourced answers in seconds. The system cites the relevant document and section, so advisors can verify before acting. Access-controlled by team level.

Integrations

SharePoint, internal policy database, client file system (read-only).

45m → 30s
average retrieval time per query
3 teams
key-person dependency reduced
92%
answer accuracy in pilot testing
Logistics 50–100 employees Ongoing

Shipment status automation and quote drafting

The problem

A mid-size freight forwarder was handling 60 to 80 inbound status calls per day for shipments already tracked in their TMS. Separately, quotes were produced manually — a process taking 2 to 4 hours per quote due to the number of variables involved. Both problems were consuming capacity that should have gone to commercial growth.

What was built

Phase one: automated status communication sent to customers at each shipment milestone. Phase two (in progress): a quote drafting assistant that generates first-draft quotes from inquiry inputs, carrier rates, and routing logic. Human reviews and sends.

Integrations

TMS (custom), email platform, carrier rate database, customer portal.

−73%
inbound status calls after phase one
Phase 2
quote automation in progress
Deep-tech / Energy Founder stage Ongoing Named client

Milesell B.V. — Positioning and commercial narrative

The problem

Milesell is developing a building-integrated urban wind energy platform optimised for turbulent rooftop environments. The technology was differentiated. The problem: it read like every other deep-tech startup — technically dense, commercially unclear, and hard to evaluate for investors, partners, and early customers.

What was built

The audit identified three compounding bottlenecks in the company narrative. We reframed the positioning around one specific constraint — urban rooftop turbulence — and rebuilt the commercial narrative, web presence, and investor-facing communication around that problem.

Transparency note

Milesell B.V. is the parent company of NorthPilot. This is a real engagement — the same diagnostic process we apply to external clients, applied to our own sister company.

  • Positioning: turbine maker → urban wind platform
  • Narrative connected to Innovatiekrediet framework
  • Investor communication rebuilt
  • Web presence restructured

What we do not claim.

These are real numbers

The metrics above come from actual pilot measurements — not projections, not industry averages. We measure against a defined baseline established before any system goes live. If we cannot measure it, we do not report it.

Results vary by context

The same system in a different company will produce different results. Data quality, process maturity, and team adoption all affect outcomes. The audit exists precisely to assess what is realistic for your specific situation before any investment is made.

Start here

The audit tells you what is realistic for your business.

45 minutes. We map your operations, identify the highest-leverage opportunity, and give you an honest assessment of what AI can actually do for your specific situation.