Guangzhou Huzhi AI Technology Co., Ltd. · HZ TECH · Guangzhou
Bringing AI into real logistics workflows
We work inside real logistics operations, using a forward-deployed engineering (FDE) approach to uncover and solve business problems, then turn high-value custom solutions into standardized software products — spanning quoting, booking, orders, warehousing and enterprise management.
Source: Guangzhou Transport Development Annual Report 2025
Industry Insight
Logistics companies don't lack software — they lack connection
The real problem: software, data and people's work are cut off from one another.
Scattered demand
Customer requests are scattered across email, WeChat, Excel and phone calls
Chaotic quoting
Rate sheets come from many sources, in inconsistent formats, with frequent version changes
Fragmented booking
Airline bookings are spread across websites, client apps, email and agent interfaces
System silos
TMS, WMS, customs, finance and tracking systems don't talk to each other
Duplicate entry
Staff copy, clean and re-key data between systems over and over
Operational black box
Management has to re-process data before it can see how the business is doing
Timing
Why now
Three technology inflection points and one shift in awareness are converging on the same timeline.
AI for everyone
Since the “DeepSeek moment”, open-source models have matched closed models on many benchmarks and token costs have fallen sharply — enterprise AI went from unaffordable to affordable.
Agents that work
A new generation of models such as GPT-5.3 and Claude Opus 4.1 has matured long-horizon planning and tool use, taking agents from demos into real workflows.
Vertical deployment
Models such as Claude Fable 5, GPT-5.6 and Kimi K3 make it realistic for agents to embed reliably into a vertical industry's systems, data and collaboration ecosystem.
Demand awakens
With domestic models like Doubao and general-purpose agents like WorkBuddy now widespread, managers and frontline staff have used AI first-hand — the question has shifted from “should we use AI?” to “how do we bring AI into our own business?”
Technology is no longer the biggest barrier. The biggest barrier is now ecosystem integration and the ability to actually deploy intelligent operations inside enterprises.
Whoever has real scenarios, industry understanding and a delivery method stands on the right side of the divide — which is why HZ Tech chose the FDE path and works with leading industry customers to go deep in logistics.
Our Method
FDE goes into the field; products come out of it
Solve concrete customer problems like an FDE team; build repeatable standard products like a SaaS company.
Go into the operation
Interview frontline roles and observe real work
Break down measurable problems
Split into steps, inputs/outputs and acceptance criteria
Build a usable prototype fast
Validate the workflow quickly with AI-assisted development
Embed in the real process
Iterate on real data, keeping human approval and rollback
Extract standard capabilities and replicate
Consolidate into standard products — agentic systems, skills, MCPs — to cut deployment cost
Be the Palantir of logistics — and more than the Palantir of logistics.
Three ways we differ from Palantir:
Palantir also embeds with customers and then productizes, but its early FDE projects had unclear outcomes, high failure rates and heavy upfront investment. We start from industry scenarios with predictable results — lighter investment, faster payoff.
We focus on the logistics vertical, with deep domain understanding and clear efficiency scenarios. Logistics has high labor costs, large volumes of valuable data and big room for gains — outcomes are more predictable and solutions transfer more easily.
Beyond standardized cost and efficiency gains, we drive revenue directly through intelligent quoting, sales conversion and capacity matching, and extend AI into the physical world.
Product Matrix
Product matrix
Warehouse Arrival Analytics & Smart Queueing
Problem solved: Warehouse labor, forklifts and dock doors were assigned by gut feel; truck arrival and waiting times were invisible and never recorded; queueing relied on manual sign-in and verbal coordination.
- Records driver arrival, waiting and handling data, building a field data asset
- Analyzes arrival peaks and waiting distribution to guide labor, forklift and dock allocation
- Drivers check in by QR code using the inbound number; full status flow from arrival, queue and call to entry and no-show
- Runs standalone; can later connect to TMS, warehouse dashboards and robot scheduling
- 20%+ higher peak throughput, 30%+ lower cost
Live at a large logistics warehouse in Baiyun, Guangzhou.
General Office Agent Platform
Problem solved: Employees lack a secure, unified, easy-to-use entry point to AI agents.
- Runs in a cloud environment the company controls — open the page and it works
- Connects team workflows horizontally: multi-agent, multi-user collaboration with shared project context and company knowledge
- Connects business workflows vertically: custom Skills and MCPs for real scenarios, integrated with WeCom, Feishu and existing business systems
- 24%–75% efficiency gains in scenarios such as dangerous-goods declarations and feasibility studies
Used by 30+ employees, including customers in other industries.
Smart Booking System
Problem solved: Repeated lookups, logins and manual entry across many airlines, websites and accounts.
- One entry point for flight, capacity and AWB status queries
- Single and batch booking task management
- Integrates via airline APIs, agent interfaces or email; uses RPA cautiously where no API exists, always with a manual fallback
- Operation logs, account permissions and exception records
AI Rate Inquiry System
Problem solved: Rates are scattered across Excel files and email attachments in inconsistent formats; staff struggle to find the latest complete, applicable price.
- Automatically ingests and parses rate files from multiple sources
- Cleans and standardizes airports, ports, airlines, currencies and units
- Detects versions and validity periods to reduce the risk of using stale rates
- Natural-language queries with traceable results — quotes only cite approved, versioned, in-date sources; missing information is routed to a person
AI-native TMS · Logistics Operations Platform
Problem solved: Order, warehouse, export, cost and business data are scattered and re-keyed by hand.
- Maintains orders, warehouses, exports, transport, costs and master data in one place
- Pulls source data from email, files and external systems to cut duplicate entry
- Management cockpit with exception alerts
- Natural-language agent entry point; critical actions keep human review (human-in-the-loop)
North America Local Freight Platform
Problem solved: Local moving and small freight in North America suffer from opaque pricing and slow matching; users struggle to find reliable providers.
- Post a moving or freight job and get live bids from drivers and movers
- Compare providers by price, rating and speed
- Pay online, with order status and fulfillment tracked end to end
- Reviews and a credit system build up provider quality
A mobile product for the North American market, currently in iterative validation.
Next: warehouse vision AMR
First understand and standardize logistics workflows through software, then train vision and task models on real data, and finally connect to robotic execution. First PoC scenario: autonomous movement of standard goods between fixed points inside a closed warehouse area.
Not an existing product. After the next funding round we will bring in a robotics technical partner, integrate a mature AMR chassis first, and complete controlled validation in a real warehouse.
Concept render · CONCEPTField Proof
Products validated in real logistics operations
First Exclusive Partner
A leading integrated logistics service group in South China
Source: public information on the customer's website
Products validated in the field
We start from our partner's real freight and warehousing processes and help them transform into an AI technology company — every product is validated and iterated in a live business setting before being consolidated into a standard capability.
- Warehouse Arrival Analytics & Smart QueueingLive · Baiyun logistics warehouse, Guangzhou
- General Office Agent PlatformLive
- Smart Booking SystemPrototype iteration
- AI Rate Inquiry SystemPrototype iteration
- AI-native TMS · Logistics Operations PlatformPrototype iteration