SIMIS.AI investor pitch - NABC v4

Same-day incident packs for childcare CCTV.

Find the clip. Draft the record. Keep children's video inside the centre.

14 days Parent request window creates urgency after an incident.

1

Incident happensThe centre must explain what happened, fast.

2

Staff scrub CCTV manuallyHours of footage, no incident-ready report.

3

Parents waitSlow answers turn into trust and escalation risk.

Child video should stay on the centre's own device or network, not copied into WhatsApp, Google Drive, or vendor laptops.

Photo: NinaZhang3 / Wikimedia Commons, CC BY-SA 4.0; cropped and darkened.

Need

Mandatory CCTV did not solve incident response.

Centres have the footage, but still need a fast, private way to find the moment and prepare a clear answer.

Need

Find the exact moment.

Constraint

Child video stays inside the centre.

Validation

8 B2B interviews narrowed SIMIS to one urgent job.

Not generic monitoring: incident search and timestamped reporting for principals.

8

Interviews

Principals feel the pain. Owners pay for risk control.

1

Wedge

Find, understand, and document the incident.

3

Proof gates

Time saved, paid LOIs, repeatable deployment.

Approach

Use existing CCTV. Add local incident intelligence.

Connect RTSP feeds, process on-prem, surface only the moments principals need.

No rip-out

Works with current cameras.

No cloud habit

Video stays on-site.

Edge AI + existing CCTV

Wedge: connect feeds, surface incidents.

Self-learning incident layer

Loop: corrections reduce false alarms.

Existing CCTV

No rip-and-replace. Connect existing RTSP feeds.

ClassroomRTSP feed

PlayroomRTSP feed

CorridorRTSP feed

OutdoorRTSP feed

→

i9 edge appliance

Local inference with children's video kept on-prem.

→

Principal screen

Incident moment + draft follow-up.

Fall risk - review Playroom A · 10:42 AM · clip + draft report ready

Feedback: principal corrects event type.

↺

Moat: site-specific learning reduces noise.

Benefits

Pilot KPI: footage hunt to same-day incident pack.

The pilot should benchmark SIMIS against each centre's manual baseline.

Clip

Event, date, camera, severity.

Record

Timestamped incident summary.

Response

Clearer parent conversation.

Manual

hours-days

SIMIS

same day

1relevant clip

1timestamped record

1parent-ready draft

Buyer workflow

Win the moment after an incident.

The principal must establish facts while staff, parents, and children are waiting. The owner pays when the workflow reduces escalation risk.

1

Incident happens

Fall, rough handling, unusual crowding, or unexplained crying.

2

AI triages

Existing RTSP camera feed is processed locally on an edge device.

3

Principal reviews

Timestamp, camera, event type, confidence, and matching clips surfaced.

4

Report drafted

Time-stamped notes reduce hearsay and manual admin.

5

Trust protected

Parent communication becomes clearer without livestreaming children.

Competition

The real competitor is manual review.

Status quo: manual review

Integrators

Preschool SaaS

SIMIS.AI

Childcare workflow

✕Manual

✕Generic

✕Admin-led

✓Purpose-built

Privacy posture

✕No controls

✕Cloud

✓Edge-first

Incident triage

✕Manual

✕Basic motion

✕No alerts

✓Alert + report

Installed CCTV fit

✕No intelligence

✕Own cams

✕Needs app

✓RTSP plug-in

Beachhead and business model

Beachhead: ~500 premium centres where trust has budget.

Singapore childcare

~1,900 centres

CCTV-installed, private sector

~1,200 centres

Beachhead: premium private

~500 centres

ICP: 6-16 cameras, 50-150 children, parent trust as a buying driver.

Pricing test

Hardware + onboarding

S$1,500

one-time pilot anchor

SaaS subscription

S$250/mo

monthly operating layer

Beachhead revenue test

S$2.25M

500 x S$4,500/yr, before partner margin

Why this fits: replaces manual investigation, not a new habit.

1

Direct pilots - 5 centres + LOIs

→

2

Installer scale - after demand proof

Use of funds and proof plan

S$150K funds 90 days to prove five paid pilot sites.

S$150K Experiment budget: edge boxes, engineering, real CCTV integrations, and paid LOIs.

5 paid pilots

Live centres, live incidents.

5 edge appliances

One box per pilot site.

2 engineers

90-day build and deployment sprint.

Days 1-30

Install boxes. Connect RTSP feeds. Capture manual baseline.

Days 31-60

Measure time-to-clip and report quality in real use.

Days 61-90

Convert demand into paid LOIs and installer handoff.

time saved

same-day report

paid LOIs

installer playbook

The ask

Raise S$150K to earn the next cheque with real centre proof.

Milestone: 5 paid pilots, faster time-to-clip, same-day reports, repeatable deployment.

Unlock: S$1-2M pre-seed to scale to 50 centres through integrator channels.

Founder fit: Singapore-Shenzhen operating access, warm childcare partner interest, and proven on-prem hardware-to-AI deployment experience.

1

Real-world proof

Deep tech needs real centres, not only a polished demo.

2

Regulatory tailwinds

Mandatory CCTV gives the market built-in infrastructure and urgency.

3

Edge moat

Privacy-first on-prem AI is the moat, not a compromise.

4

Channel scale

Principal demand first, installer channel second.