Shipsy Platform Technology and Background Intellectual Property

Shipsy develops and operates a cloud-based logistics SaaS platform, including associated AI-enabled capabilities, analytics, integration tooling, and operational methodologies. This document provides a high-level overview of the platform technologies and intellectual property that underpin Shipsy’s products and are maintained as reusable, customer-agnostic capabilities.

These assets are created and evolved as part of Shipsy’s core product and engineering roadmap, independent of any single customer deployment. They include software, architectures, data models, configuration and optimization logic, AI/ML frameworks and pipelines, reusable templates and prompts, documentation, and implementation know-how.

The technologies described here are owned by Shipsy, or used by Shipsy under applicable third-party licenses (for example, open-source components and cloud services). Shipsy continuously enhances and refines these capabilities over time and may apply them across different customers, industries, and geographies.

1.1 Core AI Agents Platform & Framework

Shipsy’s AI Agents platform and underlying agent framework used to deliver the Address Intelligence Agent and the Customer Experience (Voice) Agent.
Core platform services, APIs, orchestration layers, and runtime components that support AI agents across deployments.
Built-in multilingual support for agents (English, Arabic, and other languages) as part of Shipsy’s reusable agent stack.
Generic agent capabilities such as:

■ Session management for conversations.

■ State tracking and context management.

■ Retry / escalation logic engines (as generic building blocks).

■ Pluggable tools/adapters to call external systems from agents.

1.2 Address Intelligence Engine – Core Capabilities

Shipsy’s Address Intelligence capability is built on reusable platform components, including:

Unified address data model & store: hierarchical structures (region → city → locality → building), entity relationships, versioning, and audit logging mechanisms for address updates.
Geocoding & database referencing logic:

■ Hybrid geocoding methods using postal data, maps, GPS and other sources.

■ Algorithms for auto‑correcting spatial inaccuracies and continuously improving coordinates.

Self‑learning feedback engine framework:

■ Generic pipelines to ingest historical delivery data, driver feedback and customer edits.

■ Feedback processing logic, rules, and ML components that convert these signals into address corrections or scoring updates.

Operational intelligence capabilities:

■ Generic clustering algorithms for grouping addresses, identifying delivery “hotspots”, and supporting route optimization.

■ Stop‑level insight computation logic which works on any structured address/lat‑long dataset.

LLM / AI‑powered parsing engine:

■ Model integration layers (including where third-party models are used under license), prompt libraries, and pipelines.

■ Transliteration logic and multilingual normalization used across markets

1.3 Customer Experience / Voice Agent Platform

Shipsy’s customer experience and voice agent capabilities are delivered through reusable platform components, including:

Core conversational engine:

■ Generic dialog management framework.

■ Intent recognition and slot‑filling logic independent of any particular customer’s script.

Channel orchestration:

■ Generic capabilities for making/receiving calls, sending SMS/WhatsApp, tracking consent, and logging recordings/transcripts as platform features.

Fallback & escalation logic:

■ Platform rules to decide when to hand over to a human, how to schedule callbacks, and how to handle low‑confidence AI responses.

Agent integration hooks:

■ Generic connectors to plug into address validation services (i.e., the Address Intelligence engine) and other tools in the platform.

1.4 AI / ML Infrastructure, Tooling & Pipelines

ML pipelines for:

■ Data ingestion, cleaning, feature engineering, and training used for address quality scoring, anomaly detection, etc.

■ Evaluation frameworks, A/B testing harnesses, and performance monitoring dashboards used internally across deployments.

Generic prompt‑engineering frameworks:

■ Internal libraries for prompt templating, structured multi-step prompt orchestration, and safety guardrails (e.g., fallbacks, policy and quality controls) that are used by multiple customers.

Any internal model registries, experiment tracking systems, and model selection logic powering the agents’ AI behaviours.

1.5 Platform Integration Toolkit (Generic Connectors & APIs)

Shipsy’s platform includes standard integration capabilities designed for repeatable implementations, including:

Generic REST/GraphQL APIs exposed by the platform for:

■ Address intelligence queries (validate, normalize, geocode).

■ Triggering workflows, events, and notifications.

■ Integration with generic TMS/ERP/CRM patterns (e.g., common objects like shipment, order, consignment).

Standard reusable connectors for common classes of systems (TMS, CRMs, ticketing platforms).
Internal integration/middleware utilities:

■ Message queues, adapters, ETL scripts and reusable mappings used across multiple projects as part of the integration toolkit.

1.6 Dashboards, Analytics, and Operational Tooling

Shipsy develops and maintains standard analytics and operations tooling, including:

Generic KPI dashboards and reporting templates for:

■ Address accuracy, error reduction, and feedback utilization.

■ Call containment, unreachable rates, and delivery performance improvements.

Built‑in monitoring tools for:

■ System health and performance of agents.

■ Volume metrics, queue lengths, latency tracking.

Admin/control UIs enabling:

■ Agent configuration management.

■ User/role management, access controls.

■ Generic audit logs, history views, and event tracing.

1.7 Data Models, Schemas, and Ontologies (Generic)

Shipsy’s platform is supported by reusable data models and schemas, including:

■ Entity and schema definitions for addresses, shipments, customers, delivery attempts, call logs, etc.

■ Any canonical models, ontologies, or taxonomies used to categorize exceptions, failure causes, call outcomes, etc.

■ Standard reference tables (e.g., generic country/city structures, generic error codes) used across the platform.

1.8 Product Documentation, Playbooks, and Internal Know‑How

Product documentation, technical specifications, design references, playbooks, and architecture materials maintained by Shipsy.
Generic implementation playbooks for discovery, configuration, training, and rollout of AI agents designed for reuse across deployments.
Shipsy’s accumulated know‑how, methods, and best practices for AI deployments, including:

■ How to structure and design address enrichment projects.

■ How to benchmark voice agents.

■ How to manage feedback loops and human‑in‑the‑loop processes.

1.9 Standard platform modules and ongoing evolution

Shipsy’s logistics SaaS platform comprises a suite of standard modules and capabilities developed and maintained as part of Shipsy’s core product roadmap. The modules listed below are representative of Shipsy’s platform catalogue and may be enhanced, renamed, consolidated, or expanded over time as the platform evolves.

Core Platform includes:

■ Shipment Properties and Dashboard – shipment configuration, shipment flows (including piece‑wise flow and business pickups), Consignments Dashboard, CRM Actions.

■ First‑Mile and Last‑Mile Trip Management – trip manager, trip freeze, trip allocation, manual trip creation and full trip lifecycle handling.

■ Tracking – trip tracking using data from the Shipsy mobile app.

■ Mobile Application (Shipsy Field Ops / Driver App) – all driver‑side workflows, scanning, delivery flows, POD capture, etc.

■ Setup Section – the configuration console of the Shipsy dashboard (all masters, rules, workflows, users/roles, SLAs, etc.).

■ Shipsy BI Default Dashboard – standard operational dashboards that ship with the product.

■ Finance Operations (Basic) – COD visibility and reconciliation, driver and hub reconciliation, and end‑customer invoice generation.

■ Shipsy BI Advanced – the advanced analytics/BI module and its generic reporting, dashboarding and data‑model frameworks.

■ AI based Route Optimisation – including the Routing Playground, optimisation engine and all underlying routing/constraint algorithms used to create and optimise trips.

■ Communication Engine – the generic communication framework that configures triggers (pickup, OFD, delivery, etc.), manages templates and sends automated messages to customers via WhatsApp, email or SMS.

■ Driver Payout and Incentives – all capabilities to digitise driver payout and incentive structures (including rules based on attempts, shift timings, etc.), show payouts in real time in the driver app, and merge incentives with COD reconciliation.

■ Finance Operations – Advanced – the advanced finance module that enables digital payments and reconciles cash collections via the driver app, and by definition also includes all Finance Operations (Basic) features on which it is built.

■ Geo Coding Engine – the geocoding service that converts text addresses to latitude/longitude, together with its strategies, fallback logic and any associated reference datasets.

■ Rate Contract & Invoicing – all capabilities to digitise rate contracts (customer and vendor), configure charge types, compute charges from operations data and generate invoices.

■ LIA Co‑Pilot / Incident Management – the standard incident/exception management module, including incident capture, classification, workflows, and dashboards.

■ Hub Operations & Mid‑Mile Management – all middle‑mile and hub‑level workflows (inter‑hub movements, staging, consolidation, scans, etc.) and related dashboards.

■ Business Booking Engine / Customer Portal – the customer‑facing booking portal and APIs for order creation, tracking and self‑service.

■ Any additional standard Shipsy modules that may be added to Shipsy’s product catalogue, together with their standard features as documented in the Shipsy Knowledge Base.

Shipsy may introduce additional standard modules, features, and enhancements as part of its ongoing product development. Such additions are developed as reusable platform capabilities intended for use across deployments.