Dobby uses large‑language‑model NLP to ingest and parse shipment communications from email, PDFs, and system notifications, automatically constructing a unified, searchable timeline for each shipment. The platform applies rule‑based and machine‑learning anomaly detection and provides a conversational query interface, integrating via OAuth and API connectors for logistics operations teams.
Funding
Funding not disclosed
Founders
Product
Problem
Logistics teams often rely on fragmented email threads and legacy systems to monitor shipment progress, making it difficult to detect missed steps, delays, or exceptions in real time. This manual oversight leads to higher error rates, penalties, and reduced customer satisfaction.
Solution
Dobby applies large‑language‑model NLP to ingest and interpret all inbound shipment communications across email inboxes and connected systems. It automatically consolidates relevant data into a unified, chronological timeline for each shipment and runs rule‑based and machine‑learning anomaly detection to surface potential issues before they impact delivery. Users can query the AI‑driven knowledge base in natural language to retrieve status updates, pricing history, or compliance checks without leaving their existing workflow. The platform integrates via OAuth and API connectors, requiring no additional software installations or process redesign, enabling operations teams to scale handling of higher shipment volumes while maintaining service quality.
Target Audience
Primary users are logistics operations managers, customer service teams, and pricing/sales analysts in mid‑size to large enterprises that handle high‑volume shipments across multiple communication channels.
Features
- Deep NLP engine that parses unstructured email content, PDFs, and system notifications to extract shipment identifiers, dates, and status cues
- Automated timeline construction that aggregates all touchpoints per shipment into a searchable, visual feed
- Real‑time anomaly detection using a hybrid rule‑engine and supervised learning model to flag missed steps, delays, or compliance breaches
- Conversational query interface allowing users to ask “What is the ETA for order #12345?” and receive instant, context‑aware responses
- Pre‑built connectors for major ERP/TMS platforms and OAuth‑secured inbox integration with a few clicks
- Configurable alert routing with role‑based notifications via Slack, Microsoft Teams, or email
- Audit‑ready activity logs and granular permission controls to meet governance requirements
- Scalable cloud architecture that processes millions of messages daily while maintaining low latency