Closing the Gaps in Industry: From Bureaucracy to Data-Centered Agility

In manufacturing industries one of today’s biggest challenges is closing the knowledge and data gaps between business management, procurement, production, and IT systems teams.

Traditionally, operational implementation has been driven by the type of information that needed to be managed, but with systems that are extremely costly to implement, difficult to maintain, data-redundant, and overly fragmented across applications. The result: bureaucracy, lack of continuity, and a loss of shared vision.

Today, replacing those core systems (OLTP, ERP, PLM, CRM…) is unfeasible. The real path forward is different: reusing what already exists, but in an agile, cost-effective, data-centric way.

A New Approach: From Systems to Data

Current trends show we need to give less importance to classical applications and more to ultra-fast layers of integration and analysis, where:

• Data is the center of gravity.

• AI is applied to operational management (not design or engineering).

• The key is to close gaps through small, agile developments that deliver value quickly.

This enables:

• Natural language queries.

• Self-documentation systems.

• Lightweight and flexible user interfaces.

• Projects where technical teams and business experts collaborate seamlessly.

Case Study 1: Interface Documentation Assisted by AI

One of the most critical pain points in industry is the lack of clear and updated documentation of interfaces between systems.

• The problem: Procurement, production, logistics, and quality systems exchange data through dozens of interfaces. Documentation is often outdated or missing, and knowledge sits in the heads of a few specialists.

• The consequence: Excessive dependency, costly integration projects, and limited visibility for business leaders.

How AI can help

1. Automatic inventory of APIs, logs, messages, and database links to generate an initial interface map.

2. Dynamic documentation generation where technical details are translated into business language (e.g. “The Procurement system sends the daily parts list to Production in JSON format”).

3. Continuous updates so documentation evolves with each change.

4. Natural language queries, e.g. “Which systems consume real-time production data?”, returning a clear diagram.

Benefits

• Closes a critical gap between IT and business.

• Reduces bureaucracy and dependency on individuals.

• Accelerates decisions and new integration projects.

• Builds the foundation of a more agile, data-driven enterprise.

Case Study 2: Git and Jira.  A Knowledge System with RAG

Another strategic opportunity is applying generative AI with Retrieval-Augmented Generation (RAG) on top of Git and Jira.

How it works

• Git: the system ingests code repositories, documentation, commit history, and recent changes.

• Jira: it ingests issues, user stories, tasks, comments, attachments, and workflows.

• The content is normalized, chunked into manageable fragments, and indexed using both semantic (vector-based) and keyword search.

• A user can then ask in natural language:

  • “Which commit changed the login validation?”

  • “What issues are blocking the current sprint delivery?”

  • “Who last modified the payments module?”

• The system retrieves the relevant fragments and feeds them to the generative AI, which produces a clear, contextualized answer with links back to Git commits or Jira tickets.

What is automated

• Documentation: summaries of commits, issues, and project changes.

• Traceability: linking code changes with the Jira tasks that motivated them.

• Cross-search: a single point to query both Git and Jira without switching tools.

• Smart notifications: alerts for dependencies, blockers, or critical changes.

• Automated reporting: daily or weekly project status summaries.

Benefits

• Saves time: no need to manually search across repositories and projects.

• Improves collaboration: both business and technical users can ask questions in plain language.

• Reduces risk: better visibility of dependencies between code and tasks.

• Keeps documentation alive: always up-to-date, without extra manual effort.

• Faster decision-making: managers can ask “Which tasks are blocked by code dependencies?” and get immediate answers.

Conclusion

Digital transformation in industry is not about replacing legacy systems, but about closing the knowledge and data gaps with agile, data-driven, AI-assisted solutions.

The first step may be to improve interface documentation with AI.
The second step could be applying RAG on top of operational repositories like Git and Jira.

Both approaches empower teams to collaborate better, reduce bureaucracy, and unlock faster, more informed decision-making.

The industries that succeed in combining agility, data, and applied AI for operations will be the most competitive in the years ahead.

AI for the Development and Maintenance of Information Systems: bridging ITIL and TOGAF

Governance and AI for Development and Maintenance Environments  

Today, many organizations rely on specialized companies to provide full support for the Development and Maintenance (D+M) of their Information Systems. In this context, ensuring quality, availability and continuous evolution of IT systems is an ongoing challenge.

Our proposal is to apply the AI architecture introduced in our first post, as support to this outsourcing model, in order to enhance both operational efficiency and strategic governance.

Management first: TOGAF and ITIL as reference frameworks

Any digital transformation project must be aligned with the business and operational management methodologies already established in the organization. It is not only about adopting new technologies, but doing so in a way that strengthens the existing management framework.

Two reference frameworks stand out in this regard:

• TOGAF, as the Enterprise Architecture framework, which structures business vision, data and technology architectures.

• ITIL, as the IT Service Management framework, which defines operational best practices for handling incidents, problems, changes and continual improvement.

In our case, the approach should be top-down: starting with TOGAF’s vision and architecture phases, and landing on ITIL’s operational processes that ensure value delivery to the customer.

Where to focus in TOGAF and ITIL

Although both frameworks are broad, we can identify the most relevant aspects for controlling an AI project applied to D+M of information systems:

• TOGAF

  • Phase D: Technology Architecture, where monitoring, observability and automation platforms are defined.

  • Phase C: Information Systems Architecture, concerning operational data and logs as inputs for AI.

• ITIL

  • Incident Management, to ensure fast response to service interruptions.

  • Problem Management, to analyze root causes and prevent recurrence.

  • Event Management, to monitor systems and detect anomalies in real time.

  • Capacity and Availability Management, to anticipate needs and meet SLAs.

  • Continual Improvement, to measure and optimize outcomes.

These are the processes where the integration of operational AI can make a tangible difference.

AI architecture applied to operations: Prometheus, Grafana and ELK

The following diagram illustrates how Prometheus, Grafana and ELK act as the operational backbone of our AI architecture, linking the governance layers of TOGAF and ITIL with the advanced automation capabilities of AIOps.

Once the management framework is established, we can map it to the elements of our AI architecture that provide operational support. We have selected three well-established open-source components:

• Prometheus:

  • Real-time monitoring of metrics.

  • Collects performance data from servers, applications and databases.

  • Enables threshold-based alerts and anomaly detection.

• Grafana:

  • Visualization platform that integrates metrics and logs into unified dashboards.

  • Ideal for SLA tracking, capacity KPIs and continual improvement reporting.

  • Bridges communication between IT teams and business stakeholders.

• ELK Stack (Elasticsearch, Logstash, Kibana):

  • Centralizes and structures logs from applications, databases and infrastructure.

  • Allows fast search and historical pattern analysis.

  • Facilitates incident investigation and problem management with full traceability.

Decision automation and support to D+M

The combination of these tools does not only provide visibility, but also automates IT operations:

• Immediate detection of anomalies in logs and metrics.

• Automatic alert generation in case of incidents.

• Event correlation to identify root causes.

• Dashboards to evaluate the impact of infrastructure changes.

• Historical data for capacity planning and forecasting.

Together, they create an environment where operational decisions are driven by data and intelligent automation, aligned with ITIL and TOGAF governance.

Moving towards AIOps

This approach naturally leads us to the concept of AIOps (Artificial Intelligence for IT Operations), where AI does not only collect information but also analyzes, explains and automatically suggests actions.

Prometheus, Grafana and ELK provide the technical foundation upon which more advanced AI components (LLM, RAG) can be integrated, so that systems not only detect problems, but also interpret them and recommend solutions.

Conclusion

In Development and Maintenance of Information Systems, the key is not only having the best technology, but aligning it with management methodologies that ensure order, quality and value to the customer.

By integrating our AI architecture with TOGAF and ITIL, and supporting it with open-source tools like Prometheus, Grafana and ELK, we achieve a proactive, automated and continuously improving system.

This approach turns AI into a natural ally of D+M of information systems, strengthening enterprise and operational governance, and paving the way for a full adoption of AIOps in the future.

From LLMs to RAG, leveraging the best available tools tailored for isolated enterprise environments

From LLMs to RAG, using the best available tools.

We are kicking off this blog series with the ambition of designing the perfect on-prem AI architecture for businesses.

Enterprises everywhere face the same challenge: how to harness the power of LLMs while keeping sensitive business data fully under control. Crucially, they want these benefits without compromising security.

Organizations are asking for solutions that combine powerful large language models with a controlled, trustworthy flow of high-quality data securely managed in a fully isolated environment to avoid any risk of leakage. At the same time, there is a growing preference for open-source technologies, trusted for their transparency, flexibility, and strong security track record.
It is crucial that this architecture can seamlessly integrate with the company’s existing information systems, ensuring compatibility with current identity and authorization providers. Beyond the technical solution, clients are also seeking an extended model that includes the management of the AI system’s deployment and evolution, integrated with their existing quality pipelines, along with the ability to debug and audit how context information is being incorporated and utilized within the AI system.

Our proposal is an enterprise-ready AI architecture that starts small as a prototype focused on specific business processes but is designed to grow. Each component can be replaced or upgraded over time, ensuring long-term flexibility and performance improvements without vendor lock-in.

Enterprise RAG Architecture (On-Prem, Isolated)

Our proposed architecture is modular, open-source friendly, and fully isolated from external networks. It can start as a prototype and grow into a production-grade system without vendor lock-in.

With these three layers, enterprises get a scalable, auditable, and secure RAG environment: capable of powering digital assistants, integrating with business systems, and evolving over time.

 Layer 1    AI & Retrieval

• LLM Serving (LLaMA, Mistral, etc. via vLLM/TGI/Ollama) → Natural language understanding & generation.

• Retrieval Layer (LlamaIndex / LangChain) → Orchestrates RAG workflows.

• Vector Database + Re-ranking (FAISS/Qdrant + BGE/ColBERT) → Semantic search with high accuracy.

 Layer 2    Data & Storage

• PostgreSQL → Metadata, context, audit logs.

• MinIO (S3) → Raw documents, versions, derived chunks.

• Ingestion/ETL Pipeline (Airflow/Prefect) → Parsing, chunking, embedding, indexing.

 Layer 3    Security & Operations

• Auth & Access Control (Keycloak / SSO) → Role-based security.

• Observability (Prometheus, Grafana, ELK) → Monitor performance & quality.

• Secrets & Encryption (Vault/HSM) → Protect data & credentials.

• Caching (Redis) → Faster responses, lower cost.

 

Technology Overview and Interoperability

Why this stack “clicks”: shared standards (S3 API, SQL, OIDC/OAuth2, REST/gRPC, OpenTelemetry, Prometheus metrics), rich SDKs/connectors in LlamaIndex/LangChain, and loose coupling (object store as source of truth; vector DB as an index; Postgres for control/audit). This keeps every component replaceable without breaking the whole.

Category	Component	Function	Open Source?	Interfaces & Integration
AI & Retrieval	LLaMA / Mistral	LLMs for NLU/NLG	LLaMA (community license), Mistral (Apache-2.0)	Served via vLLM/TGI/Ollama, OpenAI-style HTTP APIs
	vLLM / TGI / Ollama	High-throughput model serving	Yes (Apache-2.0 / MIT)	REST, WebSocket, OpenAI-compatible APIs
	LlamaIndex / LangChain	RAG orchestration & pipelines	Yes (OSS, MIT)	Python/JS SDKs, connectors, REST
	FAISS / Qdrant	Vector search & retrieval	Yes (MIT / Apache-2.0)	C++/Python APIs, REST/gRPC
	Re-rankers (BGE / ColBERT)	Improves retrieval precision	Yes (Apache-2.0 / MIT)	Python models, REST wrappers
Data & Storage	PostgreSQL + JSONB	Metadata, context, audit logs	Yes (PostgreSQL license)	SQL, JDBC/ODBC, logical replication
	MinIO (S3)	Object storage for documents	Yes (AGPL-3.0)	S3 API (HTTP), SDKs
	Airflow / Prefect	ETL, ingestion, scheduling	Yes (Apache-2.0)	Python DAGs/flows, REST, CLI
Security & Operations	Keycloak	Auth, SSO, RBAC	Yes (Apache-2.0)	OIDC, OAuth2, SAML
	Prometheus + Grafana	Metrics & dashboards	Yes (Apache-2.0 / AGPL-3.0 core)	Prometheus scrape, Grafana UI/API
	ELK / OpenSearch	Logs & search	ELK (SSPL/Elastic), OpenSearch (Apache-2.0)	REST/JSON, Dashboards
	OpenTelemetry	Standard for traces/metrics/logs	Yes (Apache-2.0)	OTLP (gRPC/HTTP), SDKs
	Vault / HSM	Secrets & encryption	Vault (BSL), HSM (proprietary)	REST API, PKCS#11, KMIP
	Redis / Valkey	Caching & semantic keys	Redis (RSAL), Valkey (Apache-2.0)	RESP/TCP, TLS, client SDKs

 With this foundation in place, the real question becomes: where can AI assistants deliver the most immediate value?

From Architecture to Impact: Who Benefits First

The goal of this post is to introduce our journey toward implementing a secure, enterprise-ready RAG system. This is a starting point: in the following posts, we will move from architecture to practice, exploring how AI assistants can be applied to specific business domains.

This is just the beginning. Over the coming posts, we’ll show how these assistants can be trained and deployed, turning architectural vision into measurable operational impact.

Future posts will focus on building specialized agents for areas such as:

• Procurement Assistant
  Helps teams draft, review, and manage purchase orders and supplier contracts. Can answer questions like “What are the terms of supplier X?” or “Show me all contracts expiring this quarter.”

• Inventory & Supply Chain Assistant
  Provides quick insights on stock levels, reorder points, and supply chain risks. Can suggest replenishment actions or flag unusual consumption patterns.

• Contract Compliance Assistant
  Monitors agreements and alerts users when obligations, deadlines, or renewal dates are approaching. Helps ensure compliance without manual tracking.

• Operations Dashboard Assistant
  A conversational layer over KPIs (orders processed, delivery times, costs, SLAs). Lets managers ask, “What’s the backlog in order processing today?”

• Customer Support Knowledge Assistant
  Provides employees with instant access to resolution steps for common customer or user issues, reducing response time and improving consistency.

• Training & Onboarding Assistant
  Guides new employees through internal processes and documentation, answering “how-to” questions about operational workflows.

• Financial Operations Assistant
  Supports teams by retrieving contract values, invoice statuses, or forecasting budget impacts from changes in orders or suppliers.