Region
AWS native
- eu-west-1, us-east-1, ap-southeast-1
- IAM federation via SSO
- PrivateLink ingestion endpoints
- S3-tiered cold archive
AI-ready data fabric
Your AI initiative starts with the data.
OpenBind is the IoT data fabric from Great Wall Connect — edge software, ingestion, hot and cold storage, and live dashboards — engineered to run natively on AWS and Alibaba Cloud. Built so your operational data is clean, continuous, and queryable — the foundation every AI initiative, agent, and automation depends on.
AWS native
Alibaba Cloud native
AI-ready data substrate
One identity layer
Operating in production
3+
years in production
AI-ready
data substrate
500+
devices connected
40GB
telemetry per day
AI Readiness
AI runs on data. Most data isn't ready.
Operational signals, scattered across edge devices, brokers, and warehouses, are the #1 reason AI initiatives stall. OpenBind is the data layer that closes that gap — from sensor to model.
Continuous operational signals
Every device, every metric, every region. Buffered at the edge, ingested through a single gateway, signed and replayable.
Unified, queryable fabric
Hot Kafka topics and cold TimescaleDB hypertables behind one query surface. The data your models train on is the data your dashboards run on.
Models, agents, and dashboards
Same data, three consumers. Live dashboards for humans, agents for automation, queryable lake for your ML team. No ETL glue.
Why OpenBind
AI can't run on fragmented data.
Most AI initiatives stall in the data pipeline. OpenBind unifies edge, ingestion, hot and cold storage on AWS and Alibaba — so your data is ready before your models are.
Models starve on fragmented data
Your AI pilot died in the pipeline. Edge agents, brokers, databases, dashboards, IAM — stitched together from five vendors. OpenBind ships the seams too.
China is hard
Cross-border latency, ICP record-keeping, and Alibaba-native services do not fit cleanly onto a Western-first platform — and your AI rollout cannot skip China.
AI needs clean, continuous data
Most ops teams can't deliver a stream their models can trust. OpenBind handles ingestion, retention, and governance so your team owns the model, not the plumbing.
Architecture
From edge to insight, on two clouds.
A single control plane, two regional data planes. Each cloud runs the full stack — hot Kafka, cold TimescaleDB, and a multi-year data lake — first-class, not bolted on.
01
Edge
Lightweight agent for Linux and ARM fleets.
02
Ingestion
Idempotent MQTT/HTTP gateway, autoscaled.
03
AWS
Hot Kafka, cold TimescaleDB, multi-year data lake — us-east-1 · eu-west-1 · ap-southeast-1.
04
Alibaba Cloud
Same stack in cn-shanghai · cn-beijing · cn-hongkong — ICP-aware, OSS-tiered.
05
Dashboards
Live + historical panes, embeddable, Grafana export.
06
Consumers
Web, mobile, alerting, analytics — one identity layer across both clouds.
OpenBind runs edge software on customer fleets, ingests data through MQTT and HTTP into a horizontally-scaled gateway. Each cloud region — AWS or Alibaba Cloud — runs its own hot Kafka topics, cold TimescaleDB hypertables, and multi-year data lake. Dashboards and consumers read from both regions through one identity layer.
Platform
Eight layers, one product.
Each layer is optional in principle, but they compose. You can also bring your own broker and keep the rest.
Edge software
Buffers, batches, and signs uploads at the edge. Survives network outages. Runs on ARM and x86 fleets.
Ingestion
A horizontally-scaled gateway with per-tenant rate limits, schema validation, and write-ahead logs.
Hot & cold storage
Kafka for streaming, TimescaleDB for history. Retention and tiering are policy, not plumbing.
Live & historical dashboards
First-party dashboards with both real-time and historical panes. Embed them or export to Grafana.
Cloud-native
Built on Kubernetes and 12-factor principles. Swap AWS for another cloud, or run on your own hardware, without rewriting application code.
Data lake
Multi-year archival of raw telemetry — your training set, replayable and queryable. Compressed columnar storage with a full audit trail. The data lake your future models train on.
AI-ready by design
Every layer — edge, hot, cold, lake — exposes queryable data your models, agents, and dashboards consume directly. No ETL pipelines. No glue code.
Compliance by design
MLPS 2.0 and PIPL in China, GDPR-ready in the EU. Per-namespace retention policies, full audit logs, regional status pages — the same posture across AWS and Alibaba.
AI & Automation
AI on your operational data.
Pre-built agents that run on top of your unified data layer. Anomaly detection, prediction, and remediation — without standing up a separate ML platform.
Pattern detection
Spot anomalies in real time as they emerge from your fleet.
Predictive alerts
Forecast failures before they cascade. Route to the right on-call.
Auto-remediation
Respond to known conditions automatically — throttle, restart, dispatch.
Workflow automation
Chain multi-step responses: detect → ticket → dispatch.
Global by design
Built for China. Built for everywhere else.
OpenBind ships with the same APIs in both regions. Your code does not change.
Region
Alibaba Cloud native
- cn-shanghai, cn-beijing, cn-hongkong
- ICP-aware tenant isolation
- Internal Alibaba network for ingestion
- OSS-tiered cold archive
The same APIs in both regions. Your code does not change.
Map data © Simplemaps.com (MIT)
Industries
Built for cross-border operations
OpenBind is vertical-agnostic in principle, but proven where data has to move across borders as fluently as goods do.
Smart Manufacturing & Industry 4.0
Factory-floor equipment that must simultaneously serve local operations in China and dashboards at overseas headquarters.
Logistics & mobility
Vehicles, parcels, and personnel streaming telemetry across borders into one ingestion, storage, and alerting surface.
Global supply chain
Production in one region, brand and customers in another. Connect the data without re-platforming the application layer.
Who it's for
For developers and for operators.
For developers
OpenAPI spec, idempotent writes, Terraform modules, Prometheus metrics. The kind of platform you actually want to integrate with.
POST /v1/events
x-ob-tenant: acme-logs
{
"device": "sn-0042",
"ts": 1757900000,
"metrics": { "temp_c": 21.7 }
} - OpenAPI
- Terraform
- Prometheus
For operators
One console, role- and attribute-based access, audit logs, retention policies per namespace, regional status pages, and a 99.95% managed SLA. Less pager, more product.
- One console — unified across both clouds
- Role-based access — per-tenant and per-namespace
- Attribute-based access — fine-grained scope and policy
- Audit logs — every write, every read
- Retention policies — configurable per namespace
- Regional status pages — one URL per region
- 99.95% SLA — managed, with credits
Talk to us
Make your operations AI-ready.
Tell us about your fleet size, the regions you care about, and what you are trying to do with the data. We reply within one business day.
Or write directly to info@open-bind.com .