There’s something interesting happening in the Philippines right now when it comes to large language models. It’s not loud or overly hyped, but you can see more teams quietly moving from general software work into AI – and doing it in a way that feels practical rather than experimental.
A lot of these companies aren’t trying to build the next global model from scratch. Instead, they focus on what actually matters to businesses: integrating LLMs into existing systems, shaping them around specific workflows, and making sure they behave reliably in real use. It’s a slightly different angle, and honestly, it makes the whole space feel a bit more grounded.
1. Oski Solutions
Oski Solutions works as a software development partner that builds and maintains custom systems, with a noticeable shift toward AI-driven features, including large language model integrations. Their work is not limited to standalone AI tools. Instead, they tend to embed LLM capabilities into broader platforms – things like internal dashboards, customer-facing apps, or CMS-based systems. In practice, this often looks like adding automation layers, natural language interfaces, or data interpretation features on top of existing business logic. They also provide services in the Philippines, alongside their primary focus on clients in North America and Western and Northern Europe.
The company focuses on full-cycle development, which includes planning, building, and ongoing support, but also offers team augmentation when companies need extra hands for specific parts of a project. Their stack leans heavily on .NET, Node.js, and frontend frameworks like React or Vue, with cloud environments such as Azure and AWS used for deployment. Oski Solutions works mostly with mid-size B2B companies that rely on software for daily operations – for example, e-commerce platforms or logistics systems that need better automation or reporting. Their LLM-related work fits into that same pattern: not experimental, but tied to real workflows where automation or faster decision-making actually makes a difference.
Key Highlights:
- Provides LLM integration as part of broader software systems rather than standalone tools
- Works with clients in North America and Europe, while also offering services in the Philippines
- Focus on mid-size companies and growth-stage startups
- Combines full-cycle development with team augmentation and outsourcing
- Experience with industries like e-commerce, healthcare, fintech, and logistics
- Uses cloud-based infrastructure with Azure and AWS
Services:
- LLM services
- Custom software development
- LLM integration and AI-driven features
- Web and frontend development
- CMS development using Umbraco and WordPress
- Cloud and DevOps setup
- API development and third-party integrations
- Dedicated development teams and IT outsourcing
Contact Information:
- Website: oski.site
- E-mail: contact@oski.site
- LinkedIn: www.linkedin.com/company/oski-solutions
- Address: Kaupmehe tn 7, 10114 Tallinn, Estonia
- Phone: +48571282759
2. Global BrainForce
Global BrainForce focuses on building dedicated AI and engineering teams rather than delivering one-off projects. Their approach to large language models sits inside that model – companies don’t just get an LLM feature, they get a full team that works on it long term. That often means engineers handling everything from early experiments with language models to production systems that need to run reliably day to day.
What stands out is how much of the work is structured around stability and continuity. Global BrainForce handles the operational side – hiring, equipment, compliance – so their teams can plug into an existing product without much friction. It’s not unusual for them to support things like LLM-powered workflows or data pipelines over several years rather than short delivery cycles.
Key Highlights:
- Builds dedicated AI and LLM-focused teams instead of standalone deliverables
- Covers both experimentation and production deployment of LLM systems
- Handles operational setup including HR, compliance, and infrastructure
- Works with companies that need long-term engineering support
Services:
- LLM integrations and AI workflow development
- Machine learning and NLP engineering
- Data pipeline and analytics system development
- Dedicated development teams and staff augmentation
3. Nuxify
Nuxify works more on the application side of LLMs, focusing on how language models fit into everyday products. Their work usually revolves around integrating LLM-based features into apps or websites – things like chat-based interfaces, automated content generation, or small but useful workflow automations.
Instead of building complex AI infrastructure from scratch, Nuxify tends to focus on making LLMs usable inside existing systems. For example, adding a natural language layer to a dashboard or improving how users interact with content-heavy platforms. It’s a more lightweight approach compared to full AI platforms, but in many cases that’s exactly what companies are looking for – something that works without needing a full rebuild of their tech stack.
Key Highlights:
- Focus on LLM integrations within apps and web platforms
- Works with features like automation, personalization, and content generation
Services:
- LLM-based application integrations
- AI-driven content generation features
- Workflow automation using language models
- NLP-based user experience improvements
4. EACOMM
EACOMM approaches large language models from a slightly different angle, with a strong focus on customer service data and analytics. Their work centers around processing conversations – calls, messages, and other interactions – and turning them into something structured and usable. They provide these services often working with telecom and customer support environments where large volumes of data need to be handled consistently.
Their platform combines LLMs with speech and text processing tools to analyze interactions in detail. This includes transcription, intent detection, and sentiment analysis, along with filtering sensitive information like personal data. It’s the kind of setup that tends to sit quietly in the background but becomes important once teams start relying on it for reporting or performance tracking.
Key Highlights:
- Focus on LLM use in customer service and communication analysis
- Combines transcription, classification, and sentiment analysis
- Includes data privacy features like PII filtering
Services:
- LLM-powered customer interaction analysis
- Audio and text transcription systems
- Sentiment and intent classification
5. Lingaro Group
Lingaro Group approaches large language models from a data and operations perspective rather than treating them as standalone tools. Their work with LLMs is usually tied to improving how companies handle information – things like summarizing internal documents, pulling insights from large datasets, or helping teams search across knowledge bases without digging through multiple systems.
They also look at LLMs in combination with other technologies, particularly RPA. In practice, this means connecting language models with automated workflows – for example, extracting data from emails or invoices and passing it into structured systems. Lingaro Group tends to focus on business use cases where this combination actually replaces repetitive tasks, rather than adding another layer of complexity.
Key Highlights:
- Focus on LLM use in data processing and internal knowledge systems
- Works with document summarization and information extraction
- Explores integration between LLMs and RPA workflows
- Pays attention to adoption challenges and data readiness
Services:
- LLM-based data processing solutions
- Automation of document-heavy workflows
- Integration of LLMs with RPA systems
- Natural language search and knowledge management
6. Senti AI
Senti AI builds its work around language itself, which makes their approach to LLMs feel more grounded in real conversations than in abstract use cases. They focus a lot on how people actually communicate – across customer service channels, marketing, and internal systems – and try to make those interactions easier to manage using AI.
Their LLM-related work often shows up in conversational tools. For example, chat assistants that can switch between languages or voice systems that handle customer requests without rigid scripts. Senti AI also works on broader AI projects, but language understanding remains at the core.
Key Highlights:
- Strong focus on conversational AI and language understanding
- Developed NLP models for Filipino languages
- Works with both voice and chat-based systems
Services:
- Conversational AI and chatbot development
- LLM-based language processing
- AI strategy and consulting
- Custom AI solution development
- Managed AI services
7. Yellow.ai
Yellow.ai works on large-scale conversational systems, with LLMs playing a role in how those systems respond and adapt. Their focus is less on isolated features and more on building platforms that handle ongoing interactions – usually across customer service, sales, or support channels. In that context, LLMs are used to make conversations feel less scripted and more flexible, especially when users ask unexpected questions.
What’s noticeable is how much they emphasize automation at scale. Yellow.ai tends to position LLMs as part of a broader system that can handle a high volume of interactions without constant human input. That said, the technology is still tied to structured workflows underneath, which helps keep responses consistent. It’s the kind of setup that works best for companies dealing with large customer bases, where even small efficiency gains can add up quickly.
Key Highlights:
- Focus on conversational AI platforms at scale
- Uses LLMs to improve chatbot and voice interactions
- Works with customer service and support automation
- Combines AI responses with structured workflows
Services:
- LLM-powered chatbot development
- Conversational AI platforms
- Customer support automation
- Voice AI solutions
8. Mansoori Technologies
Mansoori Technologies designs AI systems with a strong emphasis on their performance once they are actually in use, rather than merely showcasing them in demonstrations. When they work with large language models, they typically link them to intelligent agents, information retrieval systems, and various integrations that connect directly to a business’s existing tools. They generally view these language models as just one component within a broader system, integrating them with things like APIs, databases, and established workflows, rather than expecting them to operate independently.
Furthermore, there is a clear focus on maintaining control and ensuring accountability within their systems. Mansoori Technologies dedicates effort to clearly defining the data that moves through a system, establishing how the outputs are assessed, and planning for scenarios where models might change or “drift” in their behavior over time. While they utilize tools such as LangChain, vector search, and structured evaluation pipelines, their approach to these technologies often resembles standard engineering methodologies rather than exploratory AI research.
Key Highlights:
- Focus on LLM agents, RAG systems, and workflow integrations
- Emphasis on traceability, monitoring, and controlled data access
- Works with both generative AI and classical ML when needed
Services:
- LLM agent development
- Retrieval-augmented generation (RAG) systems
- AI integration with APIs and internal tools
- NLP and text processing
9. Bots at Work
Bots at Work takes a different route compared to most LLM-focused companies. Instead of building AI systems themselves, they focus on how businesses show up inside those systems. Their work is centered around what they call AI visibility – essentially making sure a company appears in answers generated by tools like ChatGPT or other language models.
They work on structuring content, improving how information is presented, and building signals that LLMs rely on when generating answers. For example, adjusting website content so it’s easier for a model to extract direct responses, or aligning business data across platforms so it appears consistent.
Key Highlights:
- Focus on LLM visibility rather than model development
- Works with AEO, GEO, and LLM SEO concepts
- Helps businesses appear in AI-generated answers
- Connects visibility efforts with lead capture workflows
Services:
- AI visibility audits and analysis
- Content structuring for LLM extraction
- Citation and entity optimization
- LLM SEO and answer engine optimization
10. BlastAsia
BlastAsia approaches large language models through the idea of private AI systems. Instead of relying on public APIs, they focus on building internal tools that run on a company’s own data and infrastructure. Their main setup revolves around a system that works like an internal version of ChatGPT, trained or configured to respond using company-specific knowledge.
Their work with LLMs is closely tied to data control and long-term use. BlastAsia tends to position these systems as something that evolves over time – learning from interactions and adapting to specific workflows. They also combine this with more traditional software engineering, especially in environments where companies already have .NET-based systems or legacy platforms that need updating.
Key Highlights:
- Focus on private LLM implementations within company infrastructure
- Builds internal AI systems trained on business data
- Combines LLMs with existing enterprise software environments
- Works with long-term system improvement rather than quick deployments
Services:
- Private LLM development
- AI-powered internal knowledge systems
- Custom AI software development
- AI integration with business workflows
- AI transformation consulting
11. GeoPoll
GeoPoll does not build LLM systems directly, but their role sits right underneath many of them. They provide datasets that can be used to train or fine-tune language models, with a focus on real human input rather than synthetic data. In the Philippines, this includes survey responses, demographic data, and even voice recordings across multiple regions and languages.
Their work is especially relevant in cases where localization matters. For example, a chatbot trained only on global English data might miss context in Tagalog or Cebuano conversations. GeoPoll’s datasets help fill that gap by adding cultural and linguistic nuance. They also support custom data requests, which makes sense for companies building LLMs for specific industries or regions where off-the-shelf datasets fall short.
Key Highlights:
- Provides real human datasets for LLM training and fine-tuning
- Focus on survey data, demographics, and voice recordings
- Supports custom dataset creation for specific use cases
Services:
- LLM training data provision
- Custom dataset creation
- Market and demographic data collection
- Voice and text data for NLP applications
12. Accenture
Accenture assists large organizations with their broader changes involving AI, where large language models represent just one component of a much more extensive system. Their method typically involves integrating LLMs with elements such as cloud computing resources, data management platforms, and the organization’s current operational processes. They generally examine how AI can be incorporated into various operations, including customer service, supply chain management, or internal decision-making.
Practically, this often entails fitting LLMs into established workflows. For instance, they might automate common support inquiries or help teams manage substantial volumes of internal information. Accenture frequently links these systems with what they term ‘agentic AI,’ a concept where models function alongside other tools to accomplish specific tasks, rather than solely producing text.
Key Highlights:
- Works on large-scale AI and LLM integration within enterprises
- Combines LLMs with cloud, data, and operational systems
- Focus on long-term transformation rather than isolated features
- Applies LLMs in areas like customer service and supply chains
Services:
- AI and LLM integration
- Enterprise AI strategy and consulting
- Cloud and data platform development
- Automation of business processes
- AI-driven workflow optimization
13. Anycase.ai
Anycase.ai takes a very focused approach to large language models by applying them to legal research in the Philippines. Instead of building a general-purpose assistant, they’ve created a system that works specifically with legal texts – cases, statutes, and official documents.
Their use of LLMs is tightly controlled around accuracy and verification. For example, when answering a legal question, the system pulls relevant jurisprudence and explains the reasoning in a way that reflects how legal professionals actually work. It’s not trying to sound conversational for the sake of it.
Key Highlights:
- Focus on LLM use in legal research
- Built around Philippine legal sources and jurisprudence
- Provides answers with verifiable citations and references
- Designed for lawyers, law firms, and legal researchers
Services:
- LLM-powered legal research tools
- Retrieval of case law and statutes
- AI-assisted legal analysis and summaries
14. White Widget
White Widget approaches large language models as part of a broader data and software development process, rather than a standalone capability. Their work with LLMs often connects to how companies handle internal data – especially when there’s already a lot of information but no easy way to access it.
They also spend time addressing some of the known issues with LLMs, like hallucinations or unreliable outputs. White Widget looks at techniques such as retrieval-augmented generation and trust scoring to make results more transparent and easier to validate.
Key Highlights:
- Focus on LLM use for data access and natural language insights
- Works with techniques like RAG to improve accuracy
- Addresses reliability issues such as hallucinations
Services:
- LLM integration for data querying and analytics
- Retrieval-augmented generation implementation
- AI reliability and evaluation methods
- Custom software and data system development
Conclusion
If you step back and look at the full picture, LLM work in the Philippines doesn’t follow just one path. Some teams are building full systems with agents and retrieval layers, others are focused on data, and a few are solving very specific problems like legal research or customer conversations. It’s a mix, and that’s kind of the point. Instead of chasing one definition of what “AI company” should look like, most of these teams are shaping LLMs around real use cases – sometimes small, sometimes quite complex, but usually tied to something practical.
Another thing that stands out is how little of this work feels experimental for the sake of it. A lot of it is about making things work reliably – plugging LLMs into existing systems, cleaning up data, or making sure outputs can actually be trusted. In some cases, the LLM is barely visible to the end user, and that’s probably intentional. It’s just part of how the system runs now. That might not sound flashy, but it’s where most of the value seems to be building up.