DTI and Converge opened the AI and Scale-Up Center in Makati on 11 August 2026 — a free hub at the DTI Filinvest building for MSMEs to try tools, sit in training, and use workspace. That is useful. It is not a product roadmap.
This article is what to build after you visit (or after you read the press), so the trip does not become another unused login.
Coverage from Rappler, Tribune, and others describes a ~380 sqm space: coworking, demo area, training, accelerator pods, weekday hours. Converge’s pitch is connectivity and literacy. DTI’s pitch is adoption. Yours should be one operational bottleneck.
What the center is good for
Use it to:
- See tools on a decent connection instead of a sales PDF
- Get staff a shared vocabulary (what a model can and cannot do)
- Meet other operators who are also tired of spreadsheet ops
- Test whether a workflow feels useful before you pay a vendor
Do not use it as proof you are “AI-transformed.” The Tribune’s readiness vs results piece is the hangover from that claim.
What to build (the short list)
Most MSMEs we work with should build systems, then a thin AI layer — not a chatbot as the company.
1. A source of truth for the thing you sell or schedule
Bookings, orders, jobs, or inventory in one application your staff already open. If that is still Excel plus Viber, start there. Signs you have outgrown Excel and business system development are the unglamorous path.
2. One assisted workflow with a human in the loop
Examples that survive contact with a real shop:
- Draft replies using your prices and hours
- Extract invoice fields into the ledger you already trust
- Route inquiries to the right person
That is practical AI, not a demo booth.
3. Integrations you can name
If Shopee, GCash, and accounting do not talk, an AI hub will not reconcile them. Composable integrations beats five new SaaS logos from a showcase wall.
What not to build because the hub exists
| Temptation | Why to skip |
|---|---|
| “Our own ChatGPT” on the website | No inventory or policy behind it |
| Fine-tuning a model on day one | You do not have clean labeled data |
| Enterprise AI suite | You do not have a data team |
| Five pilots | None will reach production |
President-level “MSME AI” direction is policy. Your budget is still a single measured use case.
A 30-day plan after the visit
- Write the job: what takes too many minutes every day.
- Decide whether the fix is software, process, or a tool you already pay for.
- If software: scope a thin Laravel (or honest SaaS) workflow — how we scope custom work.
- Only then add AI as draft/extract.
- Keep the DTI training as literacy, not as architecture.
If you are on Laravel already, the in-app option is the AI SDK — after the workflow exists.
Straight next step
Walk-ins and slot requests go through DTI (coverage has pointed to dtistartup@dti.gov.ph / startup emails — confirm on DTI’s current page). When you know which bottleneck you want to kill, bring that one-pager to us. We will say build, buy, or wait — and we will not confuse a Makati demo with a production system.
