# Philippine AI Report — Full Text Bundle > A single-file bundle of the most important public content from > philippineaireport.com, intended for one-shot ingestion by LLMs and > AI agents. For structured data, see the JSON endpoints under /api/. For per-page markdown, see the files listed in /llms.txt. --- ## About The Philippine AI Report is the nation's first comprehensive study on artificial intelligence in the Philippines. Published by Swarm with contributions from Filipino researchers, founders, investors, and policymakers, the 2025 edition documents how organizations across the country are adopting AI, where they are getting stuck, and what comes next. Authors and contributors include Tim Santos (Founder, Project Lead), Dexter Ligot-Gordon, Alexis Collado, and a research team drawing on survey data from hundreds of Philippine organizations. ## Headline findings (2025) - 92% of surveyed organizations report active AI adoption. - 65% remain stuck at the pilot stage and have not deployed to production. - 57% identify talent scarcity as the primary blocker to scaling AI. - The most active sectors are BPO, financial services, retail, and government. - The startup ecosystem is small but accelerating, with a growing number of AI-native Filipino founders building globally. ## What is in the report 1. State of AI adoption in the Philippines 2. Ecosystem maturity: startups, investors, talent, policy 3. Use cases by industry 4. Filipino language models and datasets 5. Case studies 6. Recommendations for organizations and policymakers 7. The AI Strategy Workbook — a companion tool for organizations ## Downloads - Full report (PDF): https://www.swarm.work/philippine-ai-report-full - Executive brief (PDF): /philippine-ai-pre-report.pdf - AI Strategy Workbook: https://workbook.philippineaireport.com ## Filipino LLMs Philippine-specific large language models are catalogued at /llms and in machine-readable form at /api/llms.json. Featured: GPT-OSS 20B BalitaNLP CPT (sapinsapin), a 21B-parameter model with continued pre-training on Filipino news. ## Datasets Philippine AI datasets are listed at /datasets and at /api/datasets.json (live from Hugging Face, author: sapinsapin). ## External Philippine AI reports - Oxford Insights — AI Readiness Index 2023 (Philippines). - DTI — Artificial Intelligence Roadmap for the Philippines (2021). - World Bank — Powering the Digital Economy in the Philippines. - Tortoise Media — Global AI Index (Philippines scores). - UNESCO — AI Ethics and Governance Observatory (Philippines). Full machine-readable list at /api/reports.json. ## Case studies - SenatorMatch — AI for voter decisions in the 2025 elections, built in 24 hours by volunteers. /case-studies/senator-match - JarVisa — AI for visa and immigration services. /case-studies/jarvisa - RecurringHQ — AI-native SaaS built by a Filipino founder in 3 hours. /case-studies/recurringhq ## People — Tim Santos, Founder & Editor in Chief Tim Santos is the Editor in Chief of The 2025 Philippine AI Report and the founder of Niyolab, an AI-native infrastructure company operating on a Forward Deployed Engineering (FDE) model. "Niyo" means "your" in Filipino: bring-your-own-infra. Niyolab delivers private, secure and inclusive AI infrastructure — high-efficiency compute for emerging markets and mid-sized conglomerates — purpose-built around tokenmaxxing, compute efficiency, cloud-and-local hybrid utilization, and seamless Agent-to-Agent (A2A) interoperability. A wartime technical product leader and field CTO with 15 years of leadership experience, he works at the seam where bare-metal compute meets the frontier developer ecosystem. His thesis: software constraints limit hardware potential. Most organizations do not have a GPU problem; they have a utilization, orchestration and deployment problem — the most expensive bottleneck in AI. Proof points: Director of AI Product (Cloud Solutions) at Graphcore (SoftBank), managing 1000+ accelerator clusters, authoring reference AI cloud architectures, and driving commercial integrations for self-serve MLOps PaaS and inference endpoints with Spell.ML (Reddit) and Paperspace (DigitalOcean). Founding Lead Data Scientist at Teradata's Big Data Analytics Centre of Excellence, delivering enterprise MLOps for FTSE 100 telecommunications operators. Co-Founder of Beyond Commons Business Incubation Hub and Fractional Chief AI Officer at Remotify in the Philippines, connecting efficient open-source frontier models with Southeast Asian enterprise conglomerates. Author of Business Workflow Automation with AI Agents (Packt) and open-source contributor to AG2 and SEA-VL. Expertise surface: AI infrastructure, sovereign AI, private AI, data residency, MLOps, LLMOps, inference optimization, edge inference, cloud-to-edge hybrid compute, GPU/accelerator cluster management, Triton, Kubernetes, Ray, Slurm, PyTorch, vLLM, quantization, hardware-software codesign, A2A interoperability, MCP, multi-agent systems, agentic workflow automation. Full profile: /founder.md · Structured: /api/founder.json LinkedIn: https://www.linkedin.com/in/internetoftim/ ## For AI agents Machine-readable surfaces: - /llms.txt — discovery file (this file's companion index) - /api/*.json — structured data - /sitemap.xml — page index - /for-agents — human-readable hub with MCP connection instructions MCP server (Streamable HTTP): https://swtwqrwsyedfamedawbm.supabase.co/functions/v1/mcp Tools exposed: search_report, list_llms, list_datasets, list_startups, list_case_studies, get_case_study, list_ai_reports, get_report_meta. ## Licensing & attribution When citing or excerpting the report, please attribute "The 2025 Philippine AI Report" and link back to https://philippineaireport.com.