Perinet
Get in touch

Bootstrapped · Founded June 2026 · Türkiye

We make complex systems explorable.

Perinet builds software that turns complex systems into something people can see, change, and test. Two products today, both live: Perinet Learning for education, from primary school to university, and Perinet Mobility for public transport planning.

Live demos

See the products Talk to us

Perinet at a glance

Founded
June 2026
Stage
Bootstrapped / Early stage
Product lines
Learning · Mobility
Founder
Ahmet Ertuğrul Arık
Based in
Türkiye

01Why Perinet

Complex systems, static tools.

Digital tools often keep complex systems static: a textbook moved onto a screen, a route network drawn once and filed away. Perinet is built around a different idea: people understand more and decide better when they can change parameters, observe results, compare options, and learn through exploration. That holds for a student studying waves and for a planner redesigning a minibus network.

02Approach

One core, two products.

Learning and Mobility run on the same capabilities: interactive simulation, optimization, and visualization. New product lines grow from the same core.

  • Interactive simulation & visualization

    Virtual labs, explorable models, interactive maps, and a 3D terrain view.

    • Learning
    • Mobility

    Live

  • Optimization on real networks

    Vehicle routing with OR-Tools, 2-opt sequencing, and an experimental genetic algorithm on the real road graph.

    • Mobility

    Live

  • Terrain & energy modeling

    A direction-dependent energy model for electric minibuses on 30 m elevation data, published as open research.

    • Mobility

    Live

  • Data pipelines & open standards

    Imports KML, Shapefile, Excel, PDF, and Netcad files; exports KMZ and GTFS.

    • Mobility

    Live

  • Adaptive, AI-assisted learning

    Personalized explanations, adaptive exercises, and structured feedback.

    • Learning

    In development

  • Machine learning & deep learning

    Building on the founder's published deep learning research in geospatial applications: the next direction for the core.

    • Core

    On the horizon

03Products

Two products, live today.

Each product applies the Perinet core to one domain: scientific concepts for learners, transit networks for planners. Both have live, public demos you can open right now.

Perinet Mobility

AI-assisted public transportation planning, network analysis, and route optimization on real road networks and real terrain.

For
Municipalities, planners, and transport operators
Explore Perinet Mobility

04Perinet Learning

Learning by Seeing, Exploring, and Interacting.

From foundational learning to advanced AI concepts, Perinet Learning creates interactive experiences for different stages of education.

All three demos are publicly accessible today. Their interfaces are currently in Turkish.

Goal

Every concept in physics, chemistry, mathematics, biology, and machine learning, learnable through interactive animation.

  • Physics18 topics live · grade 11
  • Chemistry8 topics live · grade 11
  • Machine learning26 modules live
  • MathematicsSecond-grade activities live
  • BiologyPlanned

Why Perinet Learning

Built for understanding, not memorization.

  • Interactive by Design

    Students learn by manipulating, experimenting, and observing rather than only reading.

  • AI-Powered Personalization

    Learning content can adapt to the learner's age, level, and progress.

  • Visual Explanations

    Abstract concepts are transformed into animations, diagrams, simulations, and interactive models.

  • From Primary School to University

    The same learning philosophy is applied across different ages and levels of complexity.

In development

AI-Powered Learning

We are developing AI-assisted learning workflows for personalized explanations, adaptive exercises, structured feedback, interactive simulations, and guided exploration.

  1. Understand

    AI helps interpret the learner's level and learning objective.

  2. Interact

    Concepts are presented through interactive visualizations, simulations, exercises, and activities.

  3. Learn

    Students receive explanations, feedback, and progressively adapted learning experiences.

05Perinet Mobility

Public transport planning on real roads and real terrain.

An AI-assisted public transportation planning, network analysis, and route optimization platform for municipalities, planners, and transport operators.

The demo is public and available in English and Turkish. Editing and optimization run in a separate planning workspace that requires sign-in.

Screens: elevation Copernicus DEM GLO-30 (ESA/Airbus) · imagery © Esri, Maxar, Earthstar Geographics · map © OpenStreetMap contributors.

Why Perinet Mobility

Grounded in the network as it really is.

  • Network Analysis

    Lines, stops, and neighborhoods on one map, with neighborhood service coverage and per-line statistics.

  • Route Optimization

    OR-Tools vehicle routing, 2-opt sequencing, and an experimental genetic algorithm, with distance, travel time, or energy as the cost.

  • Terrain-Aware Energy

    A direction-dependent energy model for electric minibuses on 30 m elevation data, documented in an open research package (opens in a new tab).

  • Standard Exports

    Each published line downloads as KMZ for Google Earth, and the whole network as a GTFS feed, the standard format for transit data.

How it works

From raw data to a published network.

Planning happens on a revisioned draft. Nothing reaches the public map until a planner approves it and the quality checks pass.

  1. Import

    Route drawings, stop registers, and street data from KML, Shapefile, Excel, PDF, and Netcad files are merged into one road-network model.

  2. Optimize

    Optimization proposals are computed on the real road graph and compared with the current network on the map.

  3. Publish

    An approved proposal passes automatic quality checks and is published as a new release of the public map.

06Progress

Already shipping.

Not a concept: four live demos and a live pilot. We are building and testing interactive software across science, mathematics, machine learning, and public transport planning.

Counts as published on each live demo, October 2026.

  1. June 2026

    Perinet founded

    Bootstrapped, in Türkiye.

  2. October 2026 · Today

    Live and public

    • Four live demos across both product lines
    • Live pilot: the Ürgüp minibus network, 130 stops on 6 lines
    • Open research package on terrain-aware energy

  3. Next

    Growing the core

    • Every physics, chemistry, mathematics, biology, and machine learning concept as an interactive animation Goal
    • AI-assisted learning workflows In development
    • Machine learning and deep learning work On the horizon

07About

An early-stage company with live products.

Perinet is an early-stage, bootstrapped company founded in June 2026 in Türkiye. We build interactive software in two areas where complex systems are hard to grasp: education and public transportation.

Perinet Learning explores how artificial intelligence, interactive visualization, simulations, and adaptive feedback can make difficult subjects easier to understand. Perinet Mobility brings network analysis and route optimization to public transport planning, starting with the minibus network of Ürgüp, Cappadocia, built on municipal route, stop, and street data.

The same core leaves room to grow: machine learning and deep learning, building on the founder's research, is the next direction we are exploring.

Founder–product fit

  • Perinet Learning

    Education

    A degree in computer education and instructional technology, teaching in higher education since 2013, and nearly eight years coordinating a university distance-education centre.

  • Perinet Mobility

    Terrain & geospatial

    A PhD in remote sensing and GIS, with research on terrain and elevation modelling and UAV photogrammetry.

  • Next · ML & deep learning

    Deep learning research

    Published deep learning research on UAV altitude estimation, GNSS-denied localization, burned-area detection, and satellite image super-resolution.

Let's talk.

Questions, feedback, or ideas for collaboration on Perinet Learning or Perinet Mobility? Pick the conversation that fits, or write directly.

[email protected]