A KOOYA RESEARCH PROJECT
A future
in our
own words.
Philippine languages. Everyday needs.
AI with a place for every Filipino.
A future
01 / THE WAY WE SPEAK
Kumusta,
future.
Tagalog. English. Ang Taglish sa pagitan.
AI na mas malapit sa buhay natin.
02 / THE WORK STARTS HERE
Local words.
National ambition.
Kooya is adapting an open AI model for the way Filipinos speak, learn, and work.
A SMALL BEGINNING. A FILIPINO AMBITION.03 / A FUTURE WE HELP BUILD
Para sa atin.
Para sa bukas.
For us. For the future.
The next chapter is ours to build.
RESEARCH IN PROGRESS
SCROLL TO EXPLORETHE IDEA IS SIMPLE.
The future of AIshould have roomfor our voices.
Useful AI starts with understanding people. Their words. Their questions. Their everyday lives.
We’ll build on open research, test with native speakers, and let real results guide the next step.
WHAT WE’RE BUILDING
A model that
gets us.
PinoSI is Kooya’s research project to adapt an open AI model for the Philippines.
We start with an existing model and train it further on Philippine-language material. The goal: useful answers in Tagalog, English, Taglish, and more of our languages over time.
RESEARCH & DESIGN IN PROGRESSStart with a strong foundation.
We’re building on an open model, rather than training from scratch. A small first round will help us find which training approach works best.
Bring our languages into the training.
The planned mix combines Philippine-language text and code. We check licences and sources, and use native-speaker-written Taglish questions to test the result.
Make it useful. Check the work.
Coding comes first because we can run the answers. We’ll measure whether the code works, whether Taglish improves, and whether English stays strong.
Grow with evidence and partners.
Beyond the core model, we’ll explore small proof-of-concept models in other domains. Each pilot starts with one practical task, suitable data, permissions, and a clear test before it grows.
THE FIRST EXPERIMENT / PLANNED
Does local language
training make a difference?
We’ll compare two versions of the same model on the same coding problems. Only one gets the extra Philippine-language training.
A
Open model Coding training
B
Open model Philippine-language training Coding training
Same questions. Same tests. Let the results decide.
Current facts. Sources you can check.
For future advisory and public-service tools, the plan is to read approved sources when answering, explain them clearly, and cite them. Changing facts should come from current documents, with human review where needed.
THE WORK BEHIND THE WORDS
Start small.
Prove it.
Four core priorities.
Small proof-of-concept models for what comes next.
4CORE PRIORITIES
3.16BLANGUAGE TOKENS · MEASURED
COLLECTED DATA · NOT MODEL RESULTS
OUR CORE FOCUS 11 OCT 2026
Open a field to explore model-size targets and data progress.
01Regional languages & knowledgeCore focus
Our main foundation: Philippine regional languages and local knowledge. We’ll build from Tagalog, English and Taglish, with more regional languages over time.
MODEL SIZE TARGETS
RECOMMENDED V130–35B
FUTURE TARGET70B+
B = billion parameters · planned model size
3.16B tokensMeasured · including 2.37B Tagalog tokens
02STEMCore focus
Math, science, and formal reasoning. Build skills we can check through correct answers, calculations, and verifiable proofs.
MODEL SIZE TARGETS
RECOMMENDED V114–32B
FUTURE TARGET70B+
B = billion parameters · planned model size
FORMAL MATH · SPECIALIST MODEL
RECOMMENDED V17–14B
FUTURE TARGET32B
B = billion parameters · planned model size
Math~50B tokens · approximate
Science~8–12B tokens · estimated
Formal math~3–5B tokens · estimated
03CodingFirst pilot
Our first pilot: test coding answers in English and Taglish by running the code.
MODEL SIZE TARGETS
RECOMMENDED V114–32B
FUTURE TARGET70B+
B = billion parameters · planned model size
~40–70B tokensEstimated · code
04Defensive cybersecurityLightweight
A lightweight focus on defense: help explain scams, phishing, and security alerts. Keep the scope narrow, with safety review before broader work.
MODEL SIZE TARGETS
RECOMMENDED V114–32B
FUTURE TARGET70B+
B = billion parameters · planned model size
~4–7B tokensEstimated · defensive scope
SMALL MODELS. REAL-WORLD TESTS.
Prove the idea first.
For other domains, we’ll start with lightweight proof-of-concept (POC) models. One focused task. A small test. Evidence before scale.
05AgriculturePlanned POC
A small model for a focused farming question. Prove it can give a useful, checkable answer before expanding.
MODEL SIZE TARGETS
RECOMMENDED V14–8B
FUTURE TARGET14–32B
B = billion parameters · planned model size
~0.3B tokensApproximate · plus structured data
06Supply chainPlanned POC
A small model for a practical delivery, stock, or demand task. Start with a limited problem and results we can check.
MODEL SIZE TARGETS
RECOMMENDED V14–8B
FUTURE TARGET14–32B
B = billion parameters · planned model size
Structured dataNo token total
07Biosafety & preparednessPlanned POC
A small proof of concept for public-health and preparedness information. Data collection is pending.
MODEL SIZE TARGETS
RECOMMENDED V14–8B
FUTURE TARGET14–32B
B = billion parameters · planned model size
~0.1B tokensEstimated · less than 1 GB, pending
08Critical infrastructurePlanned POC
A small proof of concept using power, water, transport, or equipment simulations. Test a focused task before scaling.
MODEL SIZE TARGETS
RECOMMENDED V14–8B
FUTURE TARGET14–32B
B = billion parameters · planned model size
Tools & simulatorsNot a text corpus
Research snapshot as of 11 October 2026. Model sizes are planning targets, measured in billions of parameters. Token figures describe data, not model size or performance. Pilot models are planned; growth depends on testing and resources.
THE NEXT CHAPTER
$5MUSD · FUNDRAISING TARGET
A Philippine ambition.
A shared investment.
Explore a partnership ↗ (opens in a new tab)
US$50,000 invested by Kooya.
Our own funds. A start we believe in.
Research & designIN PROGRESS
Coding pilotFIRST PILOT PLANNED
EvaluationPLANNED
MADE HERE. FOR WHAT COMES NEXT.
Let’s build
what’s next.
Good questions. Local knowledge.
Let’s build this together.