Company

Built in 2015. Running in production in 2026.

A cross-vertical mobility platform founded on the insight that every organization moving people or goods faces the same optimization problem.

The founding insight

Three verticals. The same problem.

In 2015, three verticals surfaced the same hard problem. A few major school districts and a retailer approached us directly. A national airline came through our early partnership with Gett. All three needed complex route optimization against dozens of constraints. That was the first signal the routing problem is universal, not vertical-specific.

We built Mycelium as modular bricks any vertical could configure. The optimization engine is separated from the applications on top of it. Use the full turnkey solution, or integrate just the bricks you need via API.

Origin

Built without a sales team.

Bootstrapped, profitable, and in continuous production since 2015. 11 years of cash-flow positive operation under the founders' direction, with no investor pressure on roadmap or pricing decisions.

Customer word-of-mouth has been the only growth channel for over a decade. Every deployment was earned by the platform doing the work, not by a sales motion.

What we do

The intelligence layer between orders, fleets, carriers, and customers.

Route optimization, autonomous dispatching, dynamic pricing, white-label apps, and multi-carrier orchestration, all on one platform. The constraint library that cuts airline crew transport costs 25% (3-4 dispatchers replaced, ~3,000 rides/mo since 2017) lifts delivery efficiency 30% (a 30-vehicle fleet went from 2,400 to 1,020 drive-min/day). Improvements compound across verticals.

  • Constraint library is the moat, not the solver. Solvers are commodities. Over a decade of constraint modeling across nine verticals is not.
  • Cross-vertical intelligence. Every new rule shipped for one customer can be reused by every other customer.
  • API-first. Every UI capability is also an endpoint. Deep integration is the default, not an add-on.
  • Dynamic configuration. Rules travel per request. New customers with unique rules onboard in minutes, no release cycle required.
Architecture philosophy

Separate the engine from the apps.

Customers use the full turnkey solution, or integrate only the bricks they need via REST. The engine compounds improvements across every app above it.

APPS · 9 VERTICALS
Corporate Commute
Last-Mile Delivery
Crew Transport
Student Transport
NEMT
Field Service
On-Demand Rides
Business Park
Fuel & Oil
REST API · OAUTH2 · JWT PROJECT-SCOPED CONFIG
ENGINE · ONE CORE
VRPTW SOLVER
CONSTRAINT LIBRARY
AUTO-DISPATCHER
DYNAMIC PRICING
API SURFACE
TRACKING AGGREGATION
Constraint library compounds across every app above.
How the moat compounds

Constraint work compounds across verticals.

Every constraint shipped for one vertical accumulates as latent capability for the next. A soft service-time tolerance written for airline crew transport became a primitive available to Shufersal's delivery operation. Hazmat-corridor rules first written for US heating-oil deployments later improved safe-route compliance for student transport. Per-branch rule bundles modeled for grocery delivery now enforce corporate commute policy.

Competitors are vertical specialists. Mycelium runs nine operational shapes on one platform. The depth and cross-vertical accumulation of the constraint library is what competitors cannot replicate quickly.

Compounding since 2015, across every vertical we ship into.

How deployments expand

Every deployment we kept, we grew.

Shufersal's deployment started in 2016 as the optimization API in a few main branches. Today the same API serves every branch nationally. Multi-class vehicles, per-customer windows, per-branch rule bundles travel with each request. Thousands of API requests per hour at second-scale response. 18M+ delivery routes since 2016. 230M+ orders since 2016.

Gett's relationship started in 2016. The first Gett-channel project to ship was airline crew transport in 2017. Today it covers the full stack, optimization APIs for self-service clients, end-to-end deployments for complex enterprise customers, multi-carrier orchestration across the international footprint, and white-label apps for partner operators.

The operational shape of each customer expanded layer by layer. The constraint library compounded across all of them.

Tenure as proof

The largest deployments today are the ones that started a decade ago.

Shufersal has run home-delivery routing on Mycelium since 2016. An airline customer replaced three to four manual dispatchers with our autonomous dispatcher in 2017. Both have been running on Mycelium continuously since then; the constraint library has compounded across both deployments rather than them being re-platformed.

  1. 2015
    Platform founded
    Universal-bricks vision for route optimization.
  2. 2016
    First direct production deployment
    Shufersal home delivery.
  3. 2017
    First Gett-channel production deployment
    Airline crew transport, autonomous dispatch.
  4. 2020
    ISO 27001 certified
    First certification issued. Now on the 2022 standard.
  5. 2026
    At scale
    Nine verticals. Four continents. Continuous production.
Scale

Across nine verticals. In continuous production.

7,400+
routes optimized a day · 30M+ since 2015
89K+
individual trips a day · 360M+ since 2015
9
verticals on one backbone

Multi-tenant SaaS, multi-language, multi-currency, multi-geography.

The platform you see on this site is what a decade of operational demands shapes infrastructure into.

ISO 27001 certified. Full security posture →

Send us a week of tickets. We'll model the routes.

info@mycelium.ai · +972 54 434 4443 · Available worldwide

Send us your hardest scenario Read the case studies