Blog · Enterprise AI in practice
AI in practice: What actually works in companies
We write about what we learn building and running AI systems in companies: language models and agents, architecture and integration, computer vision and automation. With examples from our projects and figures we can back up.
All articles by topic
Five Agents, One Orchestrator, 70 Percent Less Routine Work
Why an LLM is not yet an enterprise agent, how process analysis shapes the architecture and how five agents cut routine work at RPP Group.
Read more 11 min read
theBlue.ai named an OpenAI Select Partner: a milestone for enterprise AI in production
theBlue.ai has joined the OpenAI Partner Network as an OpenAI Select Partner. What the partnership means for enterprises and their AI projects.
Read more 3 min read
Local LLMs in Practice: How We Use Them at theBlue.ai
Three local LLM systems in production: model choice, costs compared with an API, quantization, MTP, MoE and when the cloud is the better option.
Read more 14 min read
MCP, Function Calling, Code Execution: How to Connect AI to Enterprise Tools
How to connect AI to CRM, ERP and other enterprise tools: five approaches, five decision criteria and the most important security risks.
Read more 11 min read
The EU AI Act: What Companies Need to Know
Deadlines, risk classes and duties under the EU AI Act after the Digital Omnibus: what applies since August 2026 and what to do now.
Read more 10 min read
AI Developers from Germany: Custom AI Solutions for Enterprises
What an AI developer from Germany does, when custom AI development pays off, how a project runs and why location and data protection matter.
Read more 6 min read
ML Con Munich: Why the Future of AI Is Spatial
Agata Chudzińska’s keynote at ML Con Munich: why World Models and Spatial Intelligence could shape the next wave of AI.
Read more 3 min read
Choosing an AI Provider for Process Automation: What to Look For
Six criteria for choosing an AI provider: approach, integration, data protection, delivery model, references and support, with questions to ask.
Read more 7 min read
The Pilot Worked, Production Didn’t: Why AI Agents Rarely Reach Operation
Why 88% of AI pilots never reach production, the four root causes behind it and how three projects made it into operation.
Read more 7 min read
The Model Is Rarely the Problem: Integrating AI into SAP, Microsoft 365 and Jira
Three integration patterns for AI in existing systems, the role of permissions and human oversight, and how to plan your project realistically.
Read more 10 min read
Why Specialized AI Systems Are the Future of Enterprise Automation
Where generic AI platforms are enough, where they hit limits in production and what specialized AI systems deliver for enterprise automation.
Read more 5 min read
Multi-Agent Systems: Architecture, Integration, and Governance
How multi-agent systems are designed, integrated into enterprise systems and governed: protocols, architecture patterns, governance and costs.
Read more 7 min read
AI Agents 2026: From Chatbots to Autonomous Digital Employees
Where AI agents work in companies in 2026, which technology matters, where their limits are and why integration decides whether they succeed.
Read more 7 min read
When Information Pressure Rises, an AI Agent Creates Order
How an AI agent at RPP Group analyses policy documents, prepares drafts and cuts the team’s routine work by around 70%.
Read more 4 min read
Building AI that Learns and Adapts: A Case Study from MRI Diagnostics
How a neural network detects focal cortical dysplasias on MRI and why continual learning makes it more reliable, as shown in two published studies.
Read more 7 min read
The Future of Enterprise AI Starts Within Your Processes
Why enterprise AI works where it is anchored in the process, what a process analysis clarifies and how it becomes an AI system.
Read more 5 min read
Why Custom AI Often Delivers More Than an Off-the-Shelf Solution
When an off-the-shelf tool is enough and when custom AI pays off, with decision questions and three examples from our own projects.
Read more 6 min read
Custom AI Solutions for Companies: How to Successfully Begin AI Development
How an idea becomes a production AI solution: process analysis, data review, prototype, development and operation, with realistic cost ranges.
Read more 6 min read
How AI Telephony with GPT and Azure Transforms Customer Service
How a GPT-based phone assistant works, which requests it handles and why a dedicated Azure environment matters for data protection and control.
Read more 6 min read
How Companies Implement AI Successfully and Why a Clear Process Matters
Why many AI projects fail on expectations and how companies find, assess and prioritise use cases in four steps.
Read more 5 min read
How Enterprises Build Reliable RAG Systems
How retrieval-augmented generation works, which search methods exist and what successful enterprise RAG projects have in common.
Read more 5 min read
Deep Research with AI: Turning Data into Decision-Ready Knowledge
How agentic research uses a supervisor and specialised AI agents to turn internal and external sources into reliable, referenced reports for decisions.
Read more 5 min read
How RAG Automates Security and Compliance Questionnaires in MedTech
How a RAG assistant at apoQlar answers security questionnaires with sources and cuts processing time from one month to under one week.
Read more 6 min read
LLMs & AI Agents in Business Applications: Document and Data Solutions
How LLMs with tools and agents turn contracts, policies and manuals into sourced answers, and what companies need to prepare for it.
Read more 5 min read
AI Chatbots with LLMs for Businesses
How an internal AI chatbot with LLMs and several AI agents accesses company data, takes over routine work and runs securely.
Read more 7 min read
Building AI That Works: Our MVP-First Approach to Enterprise AI Workflows
How an MVP-first approach makes AI projects faster and less risky: from idea, data review and prototype to operation and monitoring.
Read more 5 min read
Deep Search vs. Deep Research: The Future of AI-Powered Information Retrieval
Deep search finds relevant content, deep research uses several AI agents to build well-sourced reports. Differences, workflow, benefits and limits.
Read more 7 min read
Why 40% of AI agent projects will be cancelled and how to avoid that outcome
Gartner expects over 40% of agentic AI projects to be cancelled by 2027. Six factors from our projects that get an AI agent into production.
Read more 6 min read
Measurable Productivity Gains with LLMs and AI Agents
What LLMs and AI agents achieve in companies, which studies document the productivity gains and how a production-ready system is built.
Read more 8 min read
LLMOps for Enterprises: Benchmarking and Comparing Vision LLMs
How companies use LLMOps to compare language and vision models, version prompts and evaluate outputs, with a workflow and tools.
Read more 5 min read
Ensuring Reliable and Scalable AI Solutions with MLOps
What MLOps is, which problems it solves, which tools it involves and how companies run machine learning models stably and at scale.
Read more 4 min read
LLMs in Visual Inspection and Quality Control
How well do GPT-4o and Gemini detect defects? Our tests on circuit boards and industrial images show where LLMs help in quality control.
Read more 5 min read
Low-Code and No-Code Platforms: A Faster Way to Build Custom Apps
What low-code and no-code platforms do, which projects they suit, where their limits are and how AI complements them.
Read more 4 min read
AI Workshops for Businesses: Why They Are a Good Choice
Why AI workshops make it easier for teams to start, which content and formats exist, and how they support AI literacy under Article 4 of the AI Act.
Read more 5 min read
Movement Recognition with AI: Use in Rehab, Sport and Workplace Safety
How AI recognises and assesses movement from camera images, where it helps in rehab, sport and workplace safety, and what matters for privacy.
Read more 4 min read
Custom LLMs: Why Tailored Language Models Matter for Businesses
Prompt engineering, RAG or fine-tuning: how companies adapt language models to their tasks, where they run them and what examples look like.
Read more 5 min read
AI Avatars for Trade Fairs, Events and Congresses
How AI avatars welcome visitors at trade fair booths and congresses and answer their questions: benefits, technology and the DCK 2024 project.
Read more 5 min read
Multimodal AI in the Enterprise: Images, Documents and Text in One System
How multimodal models process images, documents and text in one system, where they help companies and how much effort to expect.
Read more 5 min read
LLM Observability and Monitoring: Keeping AI Applications in View
What LLM observability measures, why it decides quality and cost in production, and how tools such as Langfuse and LangSmith help.
Read more 5 min read
The Benefits of AI for Businesses: What Projects Actually Measure
The benefits AI brings to companies, backed by results from projects in logistics, manufacturing, MedTech and automotive, and how to get started.
Read more 4 min read
Applications of LLM Agents Across Industries
How LLM agents are built, which tasks they take on in automotive, manufacturing, healthcare and government, and how to introduce them.
Read more 6 min read
AI Avatars in VR and Mixed Reality: What Companies Can Do Today
How AI avatars in virtual reality, mixed reality and virtual spaces support training, service and medicine, and what it takes.
Read more 4 min read
Retrieval Augmented Generation (RAG): 5 Use Cases for Companies
Five use cases for retrieval augmented generation in companies, with project examples and what matters about the data.
Read more 5 min read
AI Avatars: Bringing Digital Interactions to Life
How AI avatars welcome visitors, advise customers and answer questions: five use cases, the technology behind them and a real project.
Read more 6 min read
Zero-Shot Image Segmentation: Recognising Objects Without Large Datasets
How zero-shot models such as SAM segment objects without large training datasets, and what that means for MedTech and quality inspection.
Read more 5 min read
Spatial Computing in the Enterprise: Where It Pays Off Today
What spatial computing is, where it pays off for companies today, from training and service to MedTech, and what adoption requires.
Read more 5 min read
AI Avatars: Business Applications in Banking, HR, Sports and Sales
How companies use AI avatars in banking and finance, human resources, sports and sales. What the avatars do and what it takes to introduce them.
Read more 5 min read
Extending Microsoft 365 Copilot: How and Why
How companies connect Microsoft 365 Copilot to their own data and processes with agents and Copilot connectors, and which option fits when.
Read more 5 min read
Generative AI: 6 Use Cases for Companies
Six use cases for generative AI in companies, from summarisation and extraction to assistants, with project examples and what production requires.
Read more 5 min read
Generative AI for Summaries: Getting to the Point of Documents Faster
How generative AI summarises long documents and conversations reliably, where it pays off and how quality is kept up in production.
Read more 4 min read
Vector Databases: Data Search in the Age of Generative AI
When vector databases pay off for knowledge search, which options exist, what drives cost and how they connect to existing systems.
Read more 5 min read
AI in Medical Imaging: Where It Helps Hospitals and MedTech
Where AI relieves hospitals and MedTech companies in medical imaging, from segmentation to flagged findings, with two projects and what it takes.
Read more 4 min read
Video Analytics with AI: Where It Pays Off for Companies
Where AI video analytics pays off for companies, what it needs technically, how it runs on existing cameras and what data protection requires.
Read more 5 min read
Why Do So Many AI Projects Fail Before They Reach Production?
Why many AI projects never reach production, how they differ from IT projects and what still matters after go-live.
Read more 5 min read
theBlue.ai Joins the Microsoft AI Inner Circle Partner Program
theBlue.ai joined the Microsoft AI Inner Circle Partner Program in 2020. What it means for AI projects built on Microsoft technology.
Read more 3 min read
Predictive Maintenance: Practical Use in Companies
How predictive maintenance works, how it compares with reactive and scheduled servicing, and what companies need to start.
Read more 7 min read
How Does a Computer Know What a Text Is About? Topic Modeling for Companies
How topic modeling finds themes in thousands of documents, how LDA and newer methods work, and where companies use it.
Read more 5 min read
Tell us which process is costing you the most
Describe the process and we’ll come back within one business day with an initial assessment and a proposal for a 30-minute scoping call.
- Phase 1 is a standalone engagement, with no commitment to proceed further.
- Senior engineers from day one. No account managers, no handoffs.
- We integrate, never replace. Your systems stay exactly as they are.