Cargo-Stack reads inquiries in Chinese or English, prices them from your own ratesheets, and drafts a branded quote your team just approves — then follows the file through to booking. Built in Hong Kong for Asia’s forwarders.
Your ops team is excellent. It’s the workload that isn’t human-scale anymore.
Rates live in Excel, co-loader PDFs and old email threads. By the time the quote goes out, the first credible responder has already won the cargo.
Tariff churn, collapsing validity windows, mid-year rate spikes. A stale rate means lost margin — or a lost customer. Quoting volume keeps climbing.
Per-file TMS pricing jumped 35–50% for many forwarders in 2026 — while the quoting work stayed manual. Pay enterprise prices, still copy-paste rates at midnight.
No migration. No new system to learn. Cargo-Stack sits on top of the inbox and TMS you already use.
Gmail, Outlook or WhatsApp. Your TMS stays exactly where it is — nothing to migrate.
Lane, equipment, weight & volume, Incoterms, DG flags, ready dates — extracted from real, messy, mixed-language messages.
Excel ratesheets, co-loader PDFs, spot quotes from carriers — structured into one rate library, with your margin and validity rules applied.
A branded bilingual quote in minutes. Then validity reminders, polite chasing, win/loss capture — and a drafted booking when the customer says yes.
Pick a sample inquiry (or write your own), then run the agent. Everything below happens in your browser, on sample rates.
Demo runs fully in-browser on sample lanes (Yantian→LA, HKG→FRA, Shekou→SIN). The real product prices from your own ratesheets and knows your lanes.
Chinese and global shippers run annual tenders as hundred-lane Excel matrices with multi-round negotiation. Cargo-Stack fills them from your rate library, tells you where to fight, and never misses a deadline.
Included for founding design partners. Growth add-on at general availability.
| Lane | Volume | Target | Your bid | Suggested play |
|---|---|---|---|---|
| YTN → LAX · 40'HQ | 52 /yr | 3,800 | — | Won 3× in 2026 — bid to win |
| SHA → RTM · 40'HQ | 104 /yr | 2,950 | — | Volatile lane — hold margin |
| NGB → HAM · 40'HQ | 78 /yr | 3,050 | — | Competitive — protect 8% floor |
| SZX → SIN · LCL | 260 cbm/yr | 36 | — | Strong co-loader cost — bid |
| HKG → FRA · Air +500kg | 48 t/yr | 4.05 | — | Only with fuel clause |
| XMN → LGB · 40'HQ | 26 /yr | 2,700 | — | Below your floor — decline |
Every inquiry, quote and file the copilot touches becomes clean data. Dashboards ship built-in — and everything exports Power BI-ready for your own models.
Mixed Chinese-English threads, HK & GBA shorthand, WhatsApp and WeChat exports — parsed like a local, because it was built by one.
Every ratesheet, co-loader tariff and spot quote becomes searchable — with age warnings so nobody quotes from a dead rate.
Your rules, enforced: minimum margins, quote validity, approval thresholds, DG and oversize escalation paths.
Validity-expiry reminders, polite chases, win/loss reasons captured automatically — the CRM your team never had time to fill.
GoFreight, Magaya, CargoWise, Logitude — Cargo-Stack drafts and files alongside them. Start read-only; write-back when you trust it.
Hard tenant isolation. Your pricing never trains a competitor’s quotes. Aggregate benchmarks are opt-in only.
Transparent usage pricing that scales with the work the agent actually does — never a surprise line item.
We’re onboarding five forwarders onto their real inquiry traffic, measuring time-to-quote and win-rate together — before general availability.
No. Cargo-Stack sits on top of your inbox and works alongside your existing TMS — GoFreight, Magaya, CargoWise, Logitude or spreadsheets. It starts read-only and earns write-back. Nothing migrates, nobody re-trains.
Native. Traditional, Simplified, and the mixed 中英 shorthand real HK and GBA threads are written in — including WhatsApp/WeChat style messages, weight/volume conventions and DG references. English-only tools break on these; this is the reason Cargo-Stack exists.
Into your tenant, and nowhere else. Hard per-customer isolation; your rates and margins never influence another customer’s output. Any cross-customer benchmarking is aggregate-only and strictly opt-in.
A human approves every quote before it leaves — that’s the default, not a disclaimer. Drafts show their sources (which ratesheet, which rule) so mistakes are visible in seconds. Autonomy is earned per workflow, measured, and reversible.
Founding pilots run in Hong Kong and the GBA from Q3 2026 — five seats, real traffic, measured results. General availability follows the pilots. The design-partner form above is the way in.