Contact

Start in one of three ways.

The first step is usually a 15-minute intro call, which needs no commitment. An academic partnership and a pilot application are separate paths.
How would you like to start?

Process

  1. Step 01

    Short form

    Tell us your institution and role. The message is optional.

  2. Step 02

    15-minute call

    We listen to your prescribing and medication safety workflow and show PharmaDeux on an example case.

  3. Step 03

    Fit note

    We tell you plainly whether the product fits your institution. There is no commitment to a next step.

Email

Please do not send patient data or other special categories of personal data through our contact channels.

15-minute intro and demo

Which product?
Plain text. Please do not share patient data.0/2000

When you press Send, your request goes directly to [email protected].

Pilot protocol

Pilot protocol and LOI framework.

The Letter of Intent (LOI) planned to be signed with hospital management will use this framework for the pilot's scope, duration, data processing conditions and success criteria. This page is not binding.

Purpose and scope

Purpose
Validate the product on the institution's own data against an independent reference.
Phase A · Retrospective
Anonymised past prescription or claim data is assessed against an expert panel that has not seen the system's output.
Phase B · Shadow mode
The system assesses live data but shows nothing to physicians or specialists. Its findings are compared with the actual decisions afterwards.
Phase C · Live pilot
Optional. With ethics committee and institutional approval, alerts are shown to users on a limited ward.
Duration and wards
Agreed with the institution and written into the LOI.

Security and KVKK

Deployment
On the institution's own servers (on-prem) or in an isolated VPC dedicated to it. The pilot architecture is designed to run on-premises and in isolation. During a pilot, data is not moved outside the institution's infrastructure.
Anonymisation
Identifiers are anonymised in real time as data enters the system.
Access
Role-based. The LavieuxLabs team has no access to identifiable data.
Legal basis
Ethics committee approval, a data processing agreement and the safeguards required by Article 12 of KVKK (Turkey's data protection law).
Logging
Every assessment and every access is written to an append-only log.

Measured outcomes

Accuracy · Phase A
Sensitivity, specificity and positive predictive value.
Events detected · Phases A–B
Share of the potential adverse drug events identified by the panel that the system also detected.
Alert burden · Phase B
Alerts per 100 prescriptions.
Physician response · Phase C
Time from alert to decision, and the rate of overrides with a documented reason.
Report
Performance report shared with the institution. It also feeds the MDR clinical evaluation file.

Shield pilots measure the rejection rate, the rejection reasons caught at pre-check and specialist review time instead of adverse drug events. In no phase does the system make a decision in place of the user.

Location

The team works in Konya.

The team members are students at Konya Technical University. The address below is the faculty where we study. LavieuxLabs is not an official unit or partner of the university. If you would like to meet, please book an appointment by email first.

Meeting address

Konya Technical University

Faculty of Engineering and Natural Sciences

Akademi Mah. Yeni İstanbul Cad. No: 363/6
42250 Selçuklu / Konya, Türkiye
Get directions in Google Maps

The map is provided by Google.

Loading the map sends data such as your IP address to Google, and Google may set cookies. Details: Cookie Policy