Yingjie Liu

AI · Export Sales

Technical enough to spec it. Commercial enough to sell it.

Engineering with Business by training. I sell at the hardware end of the IT stack — I build and repair the machines myself — and I own the commercial cycle end to end, from a prospect list built from zero to payment collected. Over the past year I’ve handed the repetitive half of that to AI: buyer scoring, inbound triage, outbound cadence.

01

What I’ve built

Three projects, described by what I actually did.

01

NovocanvoExport sales & AI automation

Built the prospect list from zero — 17 companies across 8 markets in the EU and US, 10+ cold outbound campaigns, every follow-up written and owned personally. That converted into real buyer conversations: a Danish manufacturer now sampling, a German buyer in active price negotiation. Then I automated the repetitive layer — a pipeline that re-scores every buyer as new data lands, triages inbound mail and drafts the reply, so effort keeps following the accounts most likely to convert.

Role Export Sales & Business Development Scope FOB quotation across MOQ tiers (3,000–5,000 m; USD 1.30–2.40/m²) → order → shipment → payment collection Export SalesCold OutreachAI Automation

02

F1 race engineerReal-time local AI · voice

A live race engineer for F1 25 that listens, thinks and talks back — entirely on my own machine, with no cloud AI involved. Whisper (large-v3-turbo) catches the radio and is bound to a custom phrasing table so it hears F1 vocabulary correctly rather than guessing. That feeds a two-part brain: a rules layer that answers instantly (where’s my rival, yellow flag, safety car) and a local model for the hard calls (undercut, strategy change, fuel and damage trade-offs). Output goes back out through TTS, so it actually speaks. It reports rivals on track and keeps damage, positions and fuel updated — and it was trained to understand F1 and refuse to invent what it doesn’t know.

Build Whisper large-v3-turbo → custom vocabulary → rules + local SLM → TTS Training Qwen3.8-27B generated the training data and files, distilled down into a MiniCPM-2B run with reasoning off for latency Runs on RTX 4090, fully local — no cloud AI at any point Local LLMWhisperTTSF1

03

Personal AI assistantTwo personas, one local deployment

An always-on assistant I genuinely use every day, built as two switchable personas over a single memory and toolset. The Assistant side runs the day — file management, schedule checks, and the routine admin I would otherwise let slide. One click moves it to Soulmate: a warmer, conversational mode with its own tone, pacing and vocabulary. Both run on a locally deployed Qwen3.6 build, and the TTS voice was fine-tuned from real voice sources to the tone I prefer. Nothing leaves the machine.

Personas Assistant — files, scheduling, daily admin · Soulmate — conversational, different register and voice Voice TTS fine-tuned from real voice sources to match a preferred tone Runs on Qwen3.6, trained and deployed locally Local LLMTTSAgentsMemory

02

What I’m actually good at

Two halves of the same job: reading the spec sheet, and closing the deal it describes.

AI & automation

I build the automation rather than buying a demo of it.

  • Local AI end to end — Whisper speech in, TTS out, a fine-tuned small model in the middle, all on one 4090 with nothing sent to the cloud
  • Distilling a 27B model into a 2B one, so it answers inside the conversation instead of after it
  • Hybrid rules + model brains — a rules layer for instant replies, the model reserved for the genuinely hard calls
  • Email crons that go find leads on their own and draft the replies
  • Buyers re-scored continuously as new data lands — the tier list stays live, not a one-time snapshot
  • The hardware itself, by hand — desktops built from components, component-level fault isolation, overclocking

Business development

I can carry a deal from a cold email to cleared payment.

  • A 17-company prospect list across 8 markets, built from nothing
  • 10+ cold outbound campaigns with every follow-up written personally
  • FOB quotation and price negotiation across MOQ tiers
  • Order processing, shipment coordination with the factory, payment collection
  • Selling in two languages — Mandarin with China-side suppliers, English with EU and US buyers
03

About

Short, first person.

I sit on the technical side of a commercial job: I would rather read the spec sheet myself than have someone summarise it for me.

Engineering with Business at Singapore Polytechnic, then 3rd Sergeant in the Singapore Armed Forces, with an engineering internship at Capella Hotel and a short automation project at an employment agency along the way. Since 2026 I have been running export sales for a China-based production company — building the buyer list, quoting, negotiating, coordinating shipments — and rebuilding that same pipeline on AI in my own time.

Outside work: F1 and sim racing, PC builds, and any excuse to automate something repetitive.

04 · Contact

Happy to talk about export sales, local AI, or anything in between — say hello.

hello@yjqwerty.com write to me