Two raises.
One founder. One brief.
Imali raises in ZAR through Imali (Pty) Ltd. Chingu raises in KRW through a Korean entity registered on first municipal LOI. Each round is independent, each round has its own SAFE, each company can be backed without backing the other. Ebenworks the holdco is the founder’s vehicle and the brand-kit / engineering org. Not a fund of funds.
Public decks above. The full data room (cap table, financials, hiring plan, MOU) goes out on a signed mutual NDA. Ask at contact@ebstar.co.
R2-4M seed at R8-12M pre-money. SAFE.
WhatsApp business OS for SA’s 1.7M informal merchants. Live at imali.ebstar.co. 1,208 tests, ~85% coverage, ~$2/month infra cost. 18-month runway from a R3M close.
Five roles. In order. Funded by the round.
18-month hiring cost: R2.87M (without CTO) to R3.95M (with optional CTO). The first three hires close the SA-on-the-ground gap; BD opens the lender / FMCG conversation; CTO is optional and only at M6-9 if growth demands.
Senior Backend Dev · M1-2
Python / FastAPI / Postgres. R45K-65K / month. Takes infra and credit-scoring engineering off the founder’s plate.
Community Ops (SA) · M2-3
Field-team lead. Soweto, Mitchells Plain, Khayelitsha rounds. R20K-30K / month. Ideally bilingual (isiZulu or isiXhosa).
WhatsApp Dev · M3-4
Meta Cloud API, pywa, BSP integrations. R35K-55K / month.
BD / Partnerships · M4-6
Lender LOIs (Lula / Standard Bank / Merchant Capital), FMCG data partnerships, distribution partners (Vodacom Business, Flash). R40K-60K / month.
CTO (optional) · M6-9
Only if growth requires. R70K-100K / month + equity. Most rounds will not need this; the founder is full-stack.
Names, not types.
The funds and lenders we’re routing the round and the embedded-finance pilot through. Stage matters; thesis fit matters more.
₩2.5B seed at ₩5B pre-money.
Voice-AI welfare monitoring for Korea’s 2.13M elderly living alone. Backend + dashboard + family app + senior app deployed; 181 tests passing; PIPA compliance verified. KISA cert target Q3 2026. First paid municipal invoice Jan 2027.
Chingu-One-Pager.pdf · for fast skim
Chingu-Fact-Sheet.pdf · numbers in one page
Chingu-System-Presentation.pdf · 15 pages · architecture & voice stack · for the technically curious
Data-Room-Checklist.pdf · what unlocks on NDA
Why now, not later.
One law makes Chingu inevitable. The Integrated Community Care Assistance Act (지역사회통합돌봄법) took effect March 2026. It legally obligates Korean municipalities to provide regular welfare monitoring for elderly residents living alone. They cannot hire enough welfare workers to comply manually.
Korea crossed the super-aged threshold (20%+ population 65+) in December 2024, one year ahead of projection. The law arrived in March 2026. There is a 12-18 month window before KT, SKT, and Naver treat municipal eldercare AI as a strategic priority. The window is the round.
Korea-side hires on first LOI.
The first municipal LOI triggers Korean entity registration and three Korea-based hires. Until then the founder runs from Seoul, with a Korean-language DPO on retainer.
Korean DPO · on first LOI
Native Korean privacy counsel. Owns the PIPA registration, KISA cert filing, and municipal procurement docs. Retainer + part-time during pilot, full-time at first paid contract.
Voice / Dialect ML Engineer · M2-4
Native Korean. Owns Gyeongsang / Jeolla / Chungcheong corpus collection, ASR fine-tuning, dialect QA with halmonis. Native dialect coverage matters more than years of experience.
Welfare-Worker Researcher · M3-6
Eldercare or social-work background. Sits with 보건복지과 caseworkers in pilot municipalities. Owns the dashboard’s usability and the alert-routing rules.
Backend Engineer · M4-6
Self-hosted Whisper migration (Phase 2 ASR) drops COGS from ₩5K-9K/user to ₩600/user. The hire is funded by the unit-economics flip.
Korea-side capital paths.
Three capital paths. Government grants are largest by ticket but slowest; Korean VCs are fastest decision; strategic angels (KT/SKT alumni) close LOIs.
The numbers we underwrite ourselves on.
Burn is small by intent. The founder still has a day-job at Gractor (그렉터) in Korea, which runs the runway clock until full-time Q4 2026.
Chingu: not yet, by design. First paid municipal invoice Q1 2027.
Founder: Gractor salary (Korea) + Ebstar music revenue. The personal runway, not the company runway.
The six we hear most. Plain answers.
Why two companies?
Because the brief is the user, not the category. Imali and Chingu share an engineering brain, a brand kit, a writing rule set, a founder. They share nothing else. The thread is the user. The manifesto walks through this.
Solo founder. How is that not the risk?
It is the risk. Each round funds a Korea-side and a SA-side hire plan inside its first 6 months. The founder will not be solo for the duration of either SAFE. Founder bio has the day-job and full-time transition date (Q4 2026).
Why pre-seed numbers when Imali is live with 1,208 tests?
Live, no paying merchants beyond a free tier yet, no lending revenue yet. By the working definitions LPs hold us to, that’s pre-seed by capital and seed-stage by technical maturity. We’d rather underclaim and outperform.
What gives Chingu a right to win municipal procurement?
Three things: PIPA Articles 21 / 23 / 31 built in from day one (not after); KISA cert path mapped to Q3 2026; the wake-word essay (/journal) shows the register discipline that municipal procurement reads as “built by someone who has sat in a Daegu kitchen.” The Care Act tailwind (지역사회통합돌봄법, March 2026) is the demand-side push.
What if no lender signs Imali?
Imali stays a paid SaaS for SA merchants on R29 / R99 plans, not a credit-graduation business. The TAM is smaller and the upside is meaningfully lower, but the company is still a real business with real merchants. The lender is the upside; the SaaS is the floor. Break-even at ~100 paying subscribers (~1,250 users).
Are you raising Imali and Chingu together or separately?
Separately. Each company has its own SAFE, its own cap table, its own jurisdiction. An LP can back one without backing the other. Ebenworks holdco does not own the operating companies on paper. The founder does. The holdco is the brand-kit, the writing rules, and the engineering organisation that ships across both.
What we don’t have on a slide.
The same not-yet list that’s on the founder page, repeated for the partners who only read this page.
No paying merchants on Imali yet (free tier only). No paid municipal pilot for Chingu yet (first invoice Q1 2027). No KISA stamp yet (target Q3 2026). POPIA registration filed (number prints here on receipt). Korean entity registration in progress (사업자등록번호 issuing).
No co-founder on either company. The first hires (Senior Backend Dev + Community Ops on Imali; Korean DPO + Dialect ML Engineer on Chingu) are funded by the rounds. The CTO-class hire on Imali is optional and only at M6-9.
No prior fund-the-founder track record at this stage. Two prior shipped products: Imali live and instrumented; Chingu deployed and tested. Plus the academic side (MSc, Korea University; Neural Networks & IJCNN) and the music side (Ebstar, 5M+ streams), which is worth what it is and not more.
If any of those “not yet”s would be a hard no in your fund’s diligence, we’d rather know now than at term sheet.
One inbox. Read by Ebenezer.
Write to contact@ebstar.co with two lines: who you are, what stage you write cheques at, and which company (Imali / Chingu / both). We send back a 30-minute slot, the relevant deck, and a data-room link.
Public decks live on this page (linked above). The full data room (cap table, financials, hiring plan, KONEPS pilot MOU, KISA roadmap, IP register, regulatory risk register) goes out on a signed mutual NDA. The data-room checklist tells you what reviewers see once it’s signed.