This article is part of the monthly APAC AI Radar series, highlighting recent AI innovations and key company updates from various companies across the Asia Pacific. As our initiative to cover the latest AI deployments in Asia Pacific, this edition begins with Indonesia and their latest AI deployments to support business executives. Read more and uncover how Indonesia’s technology leaders aim to shape their AI initiatives and strategic plans.
| Deployment/Initiative |
| Sahabat-AI, developed by GoTo Group and Indosat Ooredoo Hutchison, is a Large Language Model (LLM) designed to enhance Indonesia’s digital sovereignty. It was relaunched in June 2025 with a new version featuring 70 billion parameters. This upgraded version includes a multilingual chatbot that can interact in Bahasa Indonesia, Javanese, Sundanese, Batak, and Balinese, reflecting Indonesia’s cultural and linguistic diversity. |
| Dr. Tania, created by Neurafarm, is an AI-powered app for smallholder farmers, using deep learning and image recognition to diagnose plant diseases and pests for 14 major crops with ~80% accuracy. It offers real-time diagnostics, personalized treatment recommendations, and a freemium model with premium features like soil testing and expert consultations. Trained on localized Indonesian datasets, it reduces chemical use, boosts yields, and enhances sustainable farming. |
| Contact Center Performance Coach powered by Bank Jago’s AI, built on Google Cloud’s Vertex AI and BigQuery, accelerates agent training by automatically analyzing customer call recordings and providing personalized feedback and targeted recommendations. It enhances agent performance, ensures consistent service quality, and reduces manual evaluation efforts, previously limited to 5% of calls. |
| Eco-Aerator by Crustea is a solar-powered aeration system for shrimp farming, using microbubbles to boost dissolved oxygen levels, enhancing shrimp health and increasing yields by up to 200%. Integrated with AI and IoT for real-time remote monitoring, it reduces energy costs by 80% compared to diesel systems and lowers carbon emissions. |
| ARVI, developed by BCA, is an AI-powered system that automates the evaluation of bank programs by analyzing operational data and generating key performance insights. Using AI algorithms and data visualization, it accelerates evaluations, reduces human error, and minimizes |
Introduction
Indonesia’s digital economy is rapidly expanding, with artificial intelligence (AI) driving significant transformation. Leading Southeast Asia in AI adoption, Indonesia sees 92% of its knowledge workers using generative AI in enterprises, according to a 2024 Microsoft and LinkedIn survey. Additionally, individual adoption is robust, with Indonesia ranking sixth globally in ChatGPT users, per a 2023 BCG survey. This dual momentum fosters innovation and positions Indonesia as a global AI hub.
One standout development is the advancement of Sahabat-AI, a joint initiative by GoTo Group and Indosat Ooredoo Hutchison. As of June 2025, the model has evolved into a 70-billion parameter large language model capable of supporting natural conversations in Bahasa Indonesia as well as five regional languages—Javanese, Sundanese, Balinese, Batak, and Indonesian. Designed with cultural relevance and digital sovereignty in mind, Sahabat-AI is now accessible through its dedicated online platform, enabling broader public engagement with locally-trained AI.
Complementing this is the ongoing development of national infrastructure, including NVIDIA’s GPU Merdeka cloud service and its planned $200 million AI center in Surakarta, which will support Indonesia’s ambition to train 20,000 AI talents and serve as a regional AI hub.
Across industries—ranging from financial services and agriculture to healthcare and aviation—AI is increasingly embedded in operational workflows, from optimizing logistics and enhancing credit scoring to enabling predictive diagnostics. However, challenges around infrastructure readiness, talent depth, and regulatory clarity still require coordinated effort across public and private sectors.To explore these advancements and their impact, Twimbit AI Radar: Indonesia Edition showcases the latest AI use cases across industries, highlighting innovations that are shaping the country’s future. This edition focuses on key sectors, including financial services, agriculture, and aviation, offering a comprehensive view of how Indonesia is leveraging AI to accelerate its digital ambitions.
Sahabat-AI: Indonesia’s First Sovereign Language Model at Scale
Sahabat-AI is a Large Language Model (LLM) developed through a national collaboration led by GoTo Group and Indosat Ooredoo Hutchison, launched officially in 2025. Trained on 70 billion parameters, it is designed to serve Indonesia’s linguistic and cultural diversity, supporting Bahasa Indonesia, Javanese, Sundanese, Balinese, and Batak.
The initiative combines contributions from a broad ecosystem: data from Kompas Gramedia, Republika, and Tempo, academic collaboration from UI, ITB, UGM, and IT PLN, and deployment on GPU Merdeka, a sovereign cloud infrastructure operated by AI Factory Lintasarta. The model is also open-source, enabling access for developers, researchers, and institutions seeking to fine-tune or build on it.
How it works
Sahabat-AI is accessible via a public chatbot interface where users can interact with the model in multiple Indonesian languages. The chatbot functions similarly to popular tools like ChatGPT and Perplexity, responding to user queries in natural language.
The model uses retrieval-augmented generation (RAG) to generate responses supported by real-time factual sources. This approach improves information reliability, particularly for context-specific or knowledge-based queries. Integration with sovereign cloud infrastructure ensures that data management, computation, and deployment remain within Indonesia’s jurisdiction.
Beyond its chatbot interface, Sahabat-AI functions as a general-purpose open-source LLM, currently in use for applications like Indira (GoPay’s virtual assistant), but available for broader use cases across sectors such as customer service, education, public services, or productivity tools.
How is it ranked?
🚀 Industry-Wide Disruption ●●●●●
Commentary
Sahabat-AI represents a significant infrastructure development in Indonesia’s AI ecosystem. As an open-source, multilingual, and sovereign LLM, it provides local developers and enterprises with a reusable foundation that reduces dependency on external models or APIs. Its public availability via a web interface increases accessibility, especially for non-English speakers.
The model’s benchmark score of 64.123 on SEA HELM (BHASA) places it ahead of several prominent regional models, including LLaMA 3.1-8B and Sea-lion v3-9B, indicating strong language performance in Southeast Asian contexts.
While it does not attempt to compete globally at the scale of frontier LLMs, Sahabat-AI contributes meaningful progress toward national AI capability-building. It establishes a shared foundation for localized AI development and sets a precedent for what sovereign AI infrastructure can look like in a developing market. As adoption and experimentation expand, its broader influence—both domestically and regionally—is likely to grow.
Dr. Tania by Neurafarm: Smarter Farming for Healthier Crops
Dr. Tania, developed by Indonesian agri-tech company Neurafarm, is an AI-powered crop protection app designed to help farmers diagnose and manage plant diseases and pests. Tailored for smallholder farmers, the app provides accessible agricultural support through a simple smartphone interface—bridging the gap where expert access is limited.
How it works
Using deep learning and image recognition, Dr. Tania analyzes photos taken by farmers to detect crop diseases and pests with around 80% accuracy across 14 major crops. Once a diagnosis is made, the app recommends targeted treatments, reducing overuse of chemicals and improving crop yields.
The platform is structured as an interactive chatbot, enabling farmers to ask questions and receive expert-level responses in a conversational format. It operates on a freemium model, offering access to over 1,400 disease management solutions, with premium features such as soil testing and personalized consultations available at a low cost.
The AI model is trained on localized datasets sourced from Indonesian farms, enhancing its relevance and accuracy for regional conditions. Beyond diagnosis, the app provides fertilizer guidance, cultivation tips, and an “ask an expert” feature that connects users with agricultural specialists. Since launch, it has facilitated over 12,000 consultations and currently supports farming activities across more than 15 hectares of land.
How is it ranked?
🌱 Strategic Innovation ●●●●○
Commentary
Dr. Tania introduces a meaningful shift in agricultural practices by embedding AI into daily farming routines. Its ability to deliver precise diagnosis, treatment advice, and farming education through a mobile app presents a clear transformation in how crop health is managed—particularly among underserved farming communities.
The integration of localized AI, low-barrier access, and sustainability features places the solution beyond incremental improvement. By reducing chemical waste by up to 15% and addressing pest- and disease-related crop losses—which contribute to nearly 40% of total crop failure in Indonesia—Dr. Tania demonstrates significant operational and economic impact.
Although its influence has yet to reshape the broader agricultural industry, the solution is already transforming core farming practices at scale. With growing adoption and an expanding knowledge base, Dr. Tania exemplifies the characteristics of Strategic Innovation within Indonesia’s agri-tech landscape.
Contact Centre Performance Coach: AI-Powered Training for Agent
Bank Jago’s AI-powered Contact Center Training Assistant is designed to accelerate the training process for customer service agents, helping them quickly meet quality assurance standards. The system enhances agent performance by providing personalized feedback and targeted recommendations, improving their skills and overall service delivery.
How it works
The Contact Center Performance Coach uses Google Cloud’s Vertex AI platform, integrating generative AI and BigQuery’s data analytics tools to analyze customer interactions. The system automatically reviews customer call recordings, assessing agent performance and identifying areas for improvement. This eliminates the need for manual evaluations, which previously covered only a small portion (5%) of calls.
The AI assistant generates personalized feedback for each agent, offering specific recommendations to enhance communication techniques, address complex queries, and improve overall performance. This automated system ensures that agents receive timely, consistent, and targeted guidance, helping them perform better while reducing the need for time-consuming manual evaluations.
How is it ranked?
🌱 Operational Transformation ●●●○○
Commentary
Bank Jago’s AI-powered Contact Center Training Assistant significantly enhances agent performance by automating feedback and analysis, reducing the training time required for agents to meet quality standards. By analyzing a larger volume of customer interactions, the system ensures more consistent and high-quality service, improving overall customer satisfaction.
The integration of Vertex AI and BigQuery enables the bank to efficiently evaluate more interactions, ensuring better service quality across the board. This innovation enhances operational efficiency by automating manual tasks, freeing up resources to handle growing customer inquiry volumes
Eco-Aerator: Revolutionizing Shrimp Farming with Solar-Powered Efficiency
How it works
What makes this system smart is its built-in sensors and AI. These track important pond conditions like oxygen, temperature, salinity, and more. When oxygen starts running low, the AI automatically turns the aerator on. Once levels are back to normal, it turns off again—saving power without sacrificing shrimp health.
Farmers don’t need to be on site to check on things. Through a connected app or dashboard, they can monitor their ponds in real time, get alerts, and even receive AI-powered suggestions on what to do next. As the system gathers more data, it continues to improve, learning how to manage each pond more efficiently over time.
How is it ranked?
🌱 Strategic Innovation ●●●●○
Commentary
Eco-Aerator brings a fresh approach to aquaculture by combining clean energy and smart automation in one easy-to-use system. It solves two big problems at once: the high cost of energy and the risk of poor pond conditions.
The use of AI to automate decisions—like when to turn the aerator on or off—means farmers no longer need to guess or rely on constant manual checks. With up to 80% savings in operational costs, and the ability to grow more shrimp with fewer resources, the impact is already significant.
While it’s still early in adoption, the combination of solar power, AI, and real-time monitoring sets a new standard for shrimp farming in Indonesia. The Eco-Aerator shows what’s possible when technology meets everyday farming, making it a clear fit for the Strategic Innovation category.
ARVI by BCA: Smarter Training Evaluation with AI Insight
ARVI—short for Artificial Intelligence and Visualization Tool—is BCA’s internal innovation to improve how training programs are evaluated. Built to address the challenges of slow, manual reviews, ARVI combines AI-powered analytics and interactive dashboards to help the bank quickly understand what’s working, what isn’t, and how future training can be improved.
How it works
Because the AI learns from patterns over time, ARVI can also surface areas of improvement that might not be obvious—like gaps in competency alignment or recurring issues across departments. Everything is delivered faster, with more accuracy, and without the need for manual number-crunching.
How is it ranked?
🌱 Operational Transformation ●●●○○
Commentary
While ARVI hasn’t yet transformed operations beyond the HR and L&D domain, it already signals how AI can bring structure, clarity, and speed to internal decision-making. As the system expands to more parts of the bank, its potential to drive wider operational improvements will only grow—earning it a place in the Operational Transformation tier.
Key Takeaway
This 8th edition of the Indonesia AI Radar highlights five recent deployments across agriculture, aquaculture, enterprise services, and language infrastructure. While the scope here is selective, it reflects a broader trend seen across past editions: AI adoption in Indonesia is deepening, diversifying, and becoming more locally grounded.
From Dr. Tania’s accessible crop diagnostics to Sahabat-AI’s sovereign language model, the use cases in this edition demonstrate how AI is moving from pilot projects to sector-level integration. Most fall under Operational Transformation or Strategic Innovation, with Sahabat-AI emerging as the first to reach Industry-Wide Disruption.
In previous Radar reports, we’ve also seen AI applied in logistics, fintech, public health, and government services—further confirming that the ecosystem is expanding beyond big tech players. The direction is clear: Indonesia is moving from AI adoption to AI infrastructure ownership.
As the ecosystem matures, the next challenges lie in scalability, inclusion, and interoperability—ensuring AI benefits not just a few use cases, but entire industries and communities.
AI Maturity Framework Introduction
The AI maturity framework used in this report offers a structured approach to evaluating the impact of AI across industries in Indonesia. It categorizes the deployment of AI solutions based on their transformative potential, ranging from basic automation to industry-wide disruption.
| Level | Description | Impact |
| Level 1: Basic Automation | Minimal impact; routine automation of repetitive tasks. | Streamlined operations with low disruption. |
| Level 2: Incremental Improvements | Modest impact with improvements in isolated areas. | Small gains in efficiency and performance. |
| Level 3: Operational Transformation | Significant improvements in workflows or resource management. | Noticeable impact on daily operations. |
| Level 4: Strategic Innovation | AI transforms core processes, delivering major efficiency gains. | High-level impact on core business functions. |
| Level 5: Industry-Wide Disruption | Disruption of industries or creation of new business models. | Major market changes, reshaping entire sectors. |
This framework helps understand how AI deployments are evolving, enabling companies to assess their readiness for digital transformation and the scale of change they can expect from AI adoption.