Case studies

Save cost, build assets, grow revenue

Three industry cases, each showing the team we put in, what we built, and how fast it paid back. All amounts in US dollars.

Discuss your use case See pricing

Case 01 · Residential rental

AI Property Concierge

Freeing property managers from answering messages so they can sell value-added services.

About 1,200 units and 4,500 tenants managed by 8 property managers — roughly 500 tenants each. Messages were missed at peak times, and a 30-minute silence could mean a tenant moving out.

Value · Cost savingLong-term rental · Services

Team we put in

Several FDE experts × 12 weeks= 120 person-days · on-site throughout, until your managers could run it themselves

What we built

Dual-channel integration: personal WeChat and WeCom
Intent recognition and response library
Auto-ticketing after 30 minutes without reply
Automatic dispatch engine
Value-added services entry point

Our price

$480KOne-off · the data asset is yours

Client benefit

$660K / yr4 manager roles saved ($520K) + $140K value-added margin

Payback

9 monthsPositive within year one

Case 02 · Law firm

Knowledge Base + Knowledge Graph

One shared memory for 45 lawyers.

45 practising lawyers across litigation, non-litigation and high-net-worth asset management. Twenty years of pleadings and research scattered across Feishu and personal laptops; new hires took 6 months to ramp up, and finding anything meant asking someone.

Value · Asset buildingThe firm owns the data and controls access

Team we put in

Several FDE experts × 12 weeks= 240 person-days · relay of industry advisor, data engineering, agent engineering and go-live support

What we built

Document cleaning and structuring
Case database
Statute – cause of action – holding knowledge graph
Three-tier permission system
Search Q&A · document drafting · tender decks

Our price

$1.08MOne-off · the data asset is yours

Client benefit

$1.44M / yr2 hrs saved per lawyer per week: 4,320 hrs/yr, about 2.4 full-time lawyers

Payback

9 monthsBuilt in the firm's own environment; it grows with use and stays when people leave

Case 03 · Equipment rental

AI Quoting & Negotiation

$4M in new annual revenue.

Annual revenue around $50M, previously only taking large contracts worth $8–9M a year. Quoting a small job by hand took two days, so sales wouldn't touch them and enquiries went to competitors. Now AI returns a price range and completes the first round of negotiation in 5 minutes — closed deals are dispatched immediately; the rest are dropped without tying up staff.

Value · Revenue growth

Team we put in

Several FDE experts × 15 weeks= 150 person-days · relay of industry advisor, data engineering, agent engineering and go-live support

What we built

Mini-program lead capture
Operations-research pricing engine: machine type / operator / fuel / distance / site conditions
5-minute AI negotiation with three-way routing (close / drop / escalate high-value to the owner)
Dispatch integration

Our price

$650KOne-off · the data asset is yours

Client benefit

$1M / yr$4M incremental revenue × 25% gross margin

Payback

8 monthsPositive within year one

One-off investment vs annual benefit (US$ thousands)

Residential rental

480

660

Law firm

1,080

1,440

Equipment rental

650

1,000

One-off investmentAnnual benefit

Payback across the three cases

8–9months

In all three cases, first-year benefit was 1.3–1.5× the investment, so each paid back within year one. From year two the asset keeps earning while the investment doesn't repeat.

Residential rental9 months
Law firm9 months
Equipment rental8 months

9 industries, all real deliveries

Machinery manufacturing

3 hours → 5 minutes

Multilingual customer-service response with no more night or holiday gaps; support team reduced by 10+ people.

Manufacturing QC

5–7 days → same day

Data from three production sites queryable the same day; abnormal batches traced in about 2 hours, with zero data leaving the company.

Engineering design

120TB · back in 1 hour

Centralised drawing management with unified permissions and backups — a full rollback within an hour after a ransomware attack.

Machinery manufacturingRetail chainsEquipment rentalManufacturing QCEngineering designBiomaterialsLegal servicesProfessional consultingAccounting & tax

All projects were delivered jointly with our ecosystem partners; client names are anonymised.

Pick any way in

Contact us

You don't need to understand AI first, or have a complete plan. Just know where it hurts most — we'll work out the rest together.

jontyhuang@gmail.com

Tell us the one process you most want fixed, and we'll get back to you quickly.

WeChat · talk to me

victor

Send me the process that hurts most, and we'll start with just that one.

HopBase platform

hop-base.com ↗

Want to integrate models directly? Find us there.

Legal entity

SingaHop Technologies Pte. Ltd.

Registered in Singapore. Ask us directly for company credentials and contracting details.

Let's start with your single most painful process — not a grand all-in-one plan.

Last decade, competitors pulled ahead with digitalisation. This decade, AI is rewriting the standards of your industry. Email us