Masli Yahaya
Pengarah Teknikal @ AINNA | CTO
30++ tahun kepakaran merentasi IT, Kejuruteraan, Automasi AI, dan E-dagang. From MEMS design to decacorn-scale systems. Contributing to AINNA's NeuralOps and autonomous operations initiatives.
Membandingkan pengulangan Ejen AI lwn Sistem Detached dengan Penghalaan Pintar untuk 100 set penyata bank PKS.
This is a concrete demonstration of the Smart Routing + Sistem Berasingan layer in the AINNA Kecekapan Flywheel. 87% is an internal benchmark on this workload.
100 PKS upload 100 different sets of bank penyata. Itu goal is to generate accurate kewangan penyata untuk all PKS.
Small & Sederhana Enterprises
Different Bank Formats
Generated Laporan
Itu Ejen AI reads, understands, classifies, calculates, validates dan generates each report individually berulang 100 times.
AI digunakan sekali untuk membina sistem detached boleh guna semula. Penyata bank dipecahkan dan dihala secara pintar dengan konteks minimum.
Fewer tokens. Same outcome. Much smarter sistem design.
This study demonstrates the Smart Routing + Sistem Berasingan layer of the Kecekapan Flywheel. 87% token reduction is an internal benchmark on the tested bank-statement workload. Full loop: Segmentation → Smart Routing → Distillation → Sistem Berasingan → Peribadi Infrastruktur. See Strategi LLM and Model Distillation.
Transparent model inputs for investor due diligence. This study supports the unit-economics framework in the pitch deck, not a company-reported revenue claim.
| Parameter | Asas Case | Kes Terbaik | Worst Case |
|---|---|---|---|
| PKS processed | 100 | 100 | 100 |
| Penyata per PKS | 12 / year | 12 / year | 12 / year |
| Avg tokens / statement (Pendekatan 1) | ~8,500 | ~6,000 | ~12,000 |
| Avg tokens / statement (Pendekatan 2) | ~1,100 | ~800 | ~1,600 |
| Token price (external API) | $0.002 / 1K | $0.0015 / 1K | $0.003 / 1K |
| Annual savings (100 PKS) | ~$5,200 | ~$7,800 | ~$3,100 |
Study date: July 2026 · Tempatan inference API fee assumed RM 0/token (GPU infrastructure amortization tracked separately)
This study is supporting evidence for the RM 2M corporate seed round. The 87% token reduction is an internal benchmark on the tested bank-statement workload, modelled with stated assumptions, not an externally audited result.
* Internal benchmark on a modelled workload. Independent audit available on investor request.
Lower GPU compute dan data pusat load
Ketara semasa gelombang haba apabila permintaan penyejukan melonjak
Structured sistem make automasi viable untuk lebih kecil businesses
Rules + selective AI + human review lapisan
Masli Yahaya
Pengarah Teknikal @ AINNA | CTO
30++ tahun kepakaran merentasi IT, Kejuruteraan, Automasi AI, dan E-dagang. From MEMS design to decacorn-scale systems. Contributing to AINNA's NeuralOps and autonomous operations initiatives.
Basic Package
Special SME traction programme by AINNA.
Every package includes a Basic AI Agent for website governance, content assistance and article management.
View SME Offer →