Bureau Veritas SMART · USPTO Patent-Pending · ESA BIC

The foundational routing
model for commercial shipping

Neurosymbolic AI that combines learned patterns with hard-coded physical and regulatory constraints. Multiple specialized agents collaborate per voyage — weather, fuel, compliance, risk, trade — producing explainable decisions for every nautical mile.

API Docs →
5 specialized agents
8 routing modes
RESTful API-first
LLM explainable

Every voyage is a constraint-satisfaction problem — weather, fuel, emissions, compliance, commercial timing. We built the system that solves all of them simultaneously, explains why, and gets better with every nautical mile.

Infrastructure

Not a routing tool. A routing model.

VesselFront is infrastructure for the next generation of maritime decision-making — designed to serve as an AI layer inside fleet management systems, foundational models, and LLM agents.

🗺️ Route Generator ✓ 3 candidates
FOC Estimator ✓ scored
🌿 Emissions Estimator ↻ evaluating
Safety Estimator queued
📈 Trade Optimizer queued

Multi-Agent System

Five specialized agents — route, fuel, emissions, safety, trade — collaborate per voyage using structured communication and game-theoretic optimization. Not a single black-box model.

ML
Predicted 6 Bf headwind zone 94% conf
RULE
Max wave constraint: 4.5m exceeded
OPT
Divert 40 nm south — Benguela current assist
Decision Route B selected · fuel saving 12.4%

Neurosymbolic Architecture

Combines learned patterns with hard-coded physical and regulatory constraints. Handles novel scenarios — new fuels, new routes — where pure ML fails. Transparent reasoning chain.

ERP
VMS
ECDIS
LLM
BI
CRM

API-First Distribution

Web-based SaaS with zero install. RESTful APIs embed into any fleet management or ERP system. Designed as an AI layer inside foundational models and LLM agents.

Route Explanation · HAIFA → RIO
Route B was selected because it avoids the 6 Bf headwind forecast for March 16–18 while exploiting the Benguela counter-current. Net fuel saving: 12.4%. CII rating maintained at A.
Why not route through Strait of Gibraltar?
Ask about this route…

LLM-Explainable Outputs

Natural-language route justifications for masters, charterers, and auditors. Audit-ready outputs for CII, EU ETS, and FuelEU Maritime compliance reporting.

Architecture

Three pillars of agentic routing

System → Distribution → User. Reimagining how navigators will chart routes tomorrow.

Pillar A — Intelligence

Neurosymbolic & Multi-Agentic AI

  • Combines learned patterns with hard-coded physical and regulatory constraints
  • Multiple specialized agents — weather, fuel, compliance, risk — collaborate per voyage
  • Handles novel scenarios (new fuels, new routes) where pure ML fails
  • Transparent reasoning chain — not a black box
Defensible architecture moat — hard to replicate
Pillar B — Distribution

API-First & Foundation-Model Ready

  • Web-based SaaS — zero install, instant fleet onboarding
  • RESTful APIs embed into any fleet management or ERP system
  • Designed to serve as an AI layer inside foundational models and LLM agents
  • Low CAC via channel partnerships
Embedded distribution — high stickiness, low churn
Pillar C — Explainability

LLM-Powered Route Intelligence

  • Natural-language route justifications for masters, charterers and auditors
  • Web + mobile delivery — decisions accessible anywhere
  • Audit-ready outputs for CII / EU ETS / FuelEU compliance
  • Positioned for emerging MASS algorithmic accountability
Explainability premium — regulatory necessity by 2030
Engine

Five agents. Eight modes. One optimized voyage.

Agents interact through structured communication protocols with game-theoretic collaboration — no hallucinations, no ambiguity.

VesselFront Route Optimization Engine · v2.4.1
All 5 agents operational
Voyage Input
vessel_imo: "9315800" from: "HAIFA" to: "RIO DE JANEIRO" depart: "2026-03-14" optimize: "fuel_efficiency" cii_target: "A" max_wave_m: 4.5
Specialized Agents
🗺️
Route Generator
Proposes candidate trajectories
FOC Estimator
Evaluates fuel cost per route
🌿
Emissions Estimator
Quantifies CO₂ per trajectory
Safety Estimator
Wind, wave, current risk
📈
Trade Optimizer
Market timing & port costs
Optimized Output
distance_nm5,847
duration_hr287.4
fuel_saving12.4%
co2_reduction_t18.7
cii_ratingA
ets_cost_saved€3,420
Route diverts 40 nm south of original rhumb line to exploit the Benguela counter-current and avoid a 6 Bf headwind zone forecast for March 16–18. Net fuel saving offsets the added distance by 2.3×.
Market

Eight drivers compounding well into the 2030s

The heart of AI in shipping is AI routing.

01

Tightening Regulation

CII ratings tighten annually. EU ETS covers shipping. FuelEU Maritime mandates escalate. Carbon border adjustments expand TAM globally.

Compliance demand compounds yearly
02

Alternative Fuels & Decarbonization

LNG, methanol, ammonia, hydrogen — sparse bunkering networks require co-optimization of route and fuel availability.

Complexity only solvable by hybrid AI
03

Insurance & Risk Pricing

P&I clubs moving to voyage-level risk pricing. Explainable route decisions equal due diligence proof and premium reductions.

Direct financial ROI per voyage
04

Real-Time Data Fusion

AIS, Copernicus satellites, port congestion, piracy feeds — API-centric architecture becomes the integration layer.

Network effects on data ingestion
05

Just-in-Time Arrival

IMO and major ports push JIT to cut anchorage emissions. Extends weather routing into full voyage optimization.

Value capture per voyage expands
06

Digital Twins & Vessel Models

Hull-specific performance models replace generic routing. Calibration to individual vessels raises switching costs.

Fleet-level lock-in and retention
07

Embedded Distribution

Fleet management platforms and maritime ERPs become distribution partners. API-first means embeddable, not standalone.

Capital-efficient growth model
08

Geopolitical Volatility

Red Sea disruptions, Arctic viability shifts, sanctions routing — dynamic re-routing is now constant, not edge-case.

Static systems become obsolete
Integration

Built for systems, not just screens

RESTful API with full OpenAPI 3.0 specification. Embed route intelligence into any fleet management or ERP system.

POST /v1/optimize cURL
curl -X POST https://api.vesselfront.com/v1/optimize \
  -H "Authorization: Bearer $VF_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "vessel_imo": "9315800",
    "from": "HAIFA",
    "to": "RIO DE JANEIRO",
    "depart": "2026-03-14",
    "mode": "fuel_efficiency",
    "constraints": {
      "cii_target": "A",
      "max_wave_m": 4.5
    }
  }'

// Response includes route waypoints,
// agent reasoning, and LLM explanation
  • Sub-second responses

    Optimization results returned in under 500ms. Webhook support for async batch processing.

  • 🔗

    Fleet management ready

    Native integration with major maritime ERPs. SDKs for JavaScript, Python, and Java.

  • 📄

    Multi-format export

    Route files in RTZ, CSV, PDF, GPX. ECDIS-compatible waypoint delivery.

  • 🔒

    Enterprise security

    OAuth 2.0 with JWT. SOC 2 compliant. Rate limiting at 100 req/min per key.

  • 🧠

    LLM-ready outputs

    Every response includes structured reasoning and natural-language explanations. Feed directly into your own AI systems.

Team

Built by maritime natives and AI researchers

A tested and harmonized team of top scientific and ethical caliber.

ST

Spero Theodorou, MD

Executive Chairman

InMode — startup to multi-billion dollar Nasdaq listing. Deep experience scaling ecosystems.

Capital Markets
CA

Constantine Abazis

Chief Executive Officer

Third-generation maritime native, 20 years in software. Knows why captains reject software — and how to build what they use.

Maritime · Product
NK

Nikos Krithariotis

Co-Founder & President

Serial entrepreneur with strategic access to major fleet owners. Drives high-level business development.

Business Dev
KK

Katerina Karabasi

CTO / VP Engineering

Expert in noisy, high-dimensional data. Translates research into production-grade code.

Engineering · ML
AA

Prof. Alexander Artikis

Chief Scientific Advisor

World authority on Complex Event Recognition. One of the most influential AI scientists globally.

AI Architecture
AK

Andreas Kouvaras

Head of Product & Ops

Maritime Law and Data Science. Ensures the AI solves real regulatory problems — EU ETS, CII.

Regulatory · Product
GK

Dr. Georgia Kalantzi

Weather Advisor

Oceanographer with global reputation. Her team delivers state-of-the-art AI wave modelling.

Oceanography · AI
MM

Michael Olivier Matt

Strategic Capital Advisor

30 years across Credit Suisse, UBS, and PENSADOR. UHNW and family office networks across Europe.

Institutional Finance
Versions