Technologies
The technologies behind practical AI and automation
Plain explanations of the technologies AARHIT Systems works with: what each one is, how it works, where it is useful, where it falls short and which of our services apply it. Written for decision-makers and technical readers alike.
How they fit together
No single technology makes a system intelligent
Real solutions combine several technologies. A document assistant may use data engineering, NLP, a large language model, retrieval and MLOps. An inspection system may use IoT, edge AI, computer vision and a digital twin. This table shows common combinations.
| Solution type | Technologies usually involved | Where to start reading |
|---|---|---|
| Knowledge assistant or copilot | Large language models, NLP, data engineering, MLOps | Generative AI |
| Governed task automation | Agentic AI, intelligent automation, LLMs | Agentic AI |
| Visual inspection | Computer vision, edge AI, IoT | Computer Vision |
| Forecasting and planning | Machine learning, data engineering, MLOps | Machine Learning |
| Connected operations | IoT, edge AI, digital twins, cloud AI | Internet of Things |
| Document processing | NLP, multimodal AI, intelligent automation | Natural Language Processing |
Explore
Filter technologies by group
Select a group to narrow the list, or open any technology for a full explanation.
- Artificial IntelligenceMethods that let software perceive, reason, predict, generate and act on information.
- Generative AIModels that produce text, images, code and other content from patterns learned in data.
- Agentic AIAI systems that plan multi-step work and use tools to complete it within set boundaries.
- Machine LearningAlgorithms that learn patterns from data to classify, predict, rank or detect anomalies.
- Large Language ModelsNeural networks trained on large text corpora that understand and generate language.
- Computer VisionTechniques that interpret images and video for detection, inspection, measurement and search.
- Natural Language ProcessingMethods for understanding, extracting from and producing human language.
- Multimodal AIModels that work across several data types such as text, images, audio and sensor streams.
- Data EngineeringThe pipelines, storage and quality controls that make data usable for analytics and AI.
- MLOpsPractices and tooling for deploying, monitoring and maintaining machine learning in production.
- Intelligent AutomationAutomation that combines workflows, bots and AI to handle judgement-heavy processes.
- Cloud AIManaged cloud services and infrastructure for building, training and serving AI.
- Internet of ThingsConnected devices and sensors that capture physical-world data and act on it.
- Edge AIRunning AI models on or near devices for low latency, privacy or limited connectivity.
- Digital TwinsVirtual representations of physical assets or processes kept in sync with real data.
Group 1
Foundations
The core disciplines that most other AI capabilities build on.
Group 2
Generative and Language
Technologies that understand, generate and act through language.
Generative AI
Models that produce text, images, code and other content from patterns learned in data.
Learn more Generative and LanguageAgentic AI
AI systems that plan multi-step work and use tools to complete it within set boundaries.
Learn more Generative and LanguageLarge Language Models
Neural networks trained on large text corpora that understand and generate language.
Learn more Generative and LanguageNatural Language Processing
Methods for understanding, extracting from and producing human language.
Learn moreGroup 3
Perception
Technologies that interpret images, video, audio and combinations of data.
Group 4
Data and Operations
The data, automation and operational practices that make AI dependable.
Data Engineering
The pipelines, storage and quality controls that make data usable for analytics and AI.
Learn more Data and OperationsMLOps
Practices and tooling for deploying, monitoring and maintaining machine learning in production.
Learn more Data and OperationsIntelligent Automation
Automation that combines workflows, bots and AI to handle judgement-heavy processes.
Learn moreGroup 5
Infrastructure and Edge
Where AI runs: cloud platforms, connected devices, the edge and digital twins.
Cloud AI
Managed cloud services and infrastructure for building, training and serving AI.
Learn more Infrastructure and EdgeInternet of Things
Connected devices and sensors that capture physical-world data and act on it.
Learn more Infrastructure and EdgeEdge AI
Running AI models on or near devices for low latency, privacy or limited connectivity.
Learn more Infrastructure and EdgeDigital Twins
Virtual representations of physical assets or processes kept in sync with real data.
Learn moreNext step
Working out which technology fits your problem?
Describe the problem rather than the technology. We will explain which approaches are realistic, what they need and how to test them.