Alleantia Agents

Build and use industrial intelligence with any LLM

Alleantia Agents brings AI agents for knowledge, data analysis and specialized tasks to the factory. They enrich machine data with your people’s experience and make it available to everyone on the shop floor.

Talk to an expert Meet the agents

New analysisAnalyst Agent
Where am I losing output on line 2?
  1. UnderstandsLine 2 · 6 machines · last 30 days
  2. Interprets38% of losses originate at press P-04
  3. ActsCreates the insight and alerts the line supervisor
Production losses · Line 2Production
Press P-0438%
Welder S221%
Robot R114%
Painting9%

Sample data

Compare press OEE across the two plants
  1. Understands12 presses · 2 plants · 6 weeks
  2. InterpretsPlant B loses 6 points of availability
  3. ActsShares the report with both plant managers
Press OEE · Plants A and BPerformance
A 81,2%B 75,4%
S1
S2
S3
S4
S5
S6

Sample data

Alert me if oven 3 goes above 240 °C
  1. UnderstandsOven 3 · zone 2 temperature probe
  2. InterpretsThreshold exceeded 4 times this week
  3. ActsActivates the rule and alerts maintenance
Oven 3 monitor · zone 2Control
240 °C
Rule active · maintenance alert

Sample data

How much energy do I use per part produced?
  1. UnderstandsLine 1 · energy meters and part counts
  2. InterpretsPeaks occur during night-shift changeovers
  3. ActsSaves the insight and replicates it on other lines
Energy per part · Line 1Energy
0,84kWh per part−6% vs. last month

Sample data

Secure by design•Grounded in data•Production-ready

Alleantia Agents works on real data from your machines and on your factory’s knowledge, with each user’s permissions: it connects data, documents and experience to handle even the most complex tasks and, when information is missing, it asks for it instead of making it up.

A single knowledge base

All your Knowledge in one platform

Documents, machine data, production events and people’s experience flow into a single knowledge graph, available to everyone on the shop floor.

Analyst AgentListening

Graph path

Sample data

12 entities · 16 relationships

  • Machines
  • Tables
  • Variables
  • Concepts
  • Documents
  • Events
  • Team notes

Every manual, one question away

Upload manuals, data sheets and procedures to the platform. The Knowledge Agent analyzes them, links them to the machines and concepts in the knowledge graph and makes them searchable in natural language.

  • RAG vector database and semantic graph, always up to date.
  • Every answer cites the documents it comes from.
Knowledge AgentPress P-04
How do I reset alarm E-217 on press P-04?
Check the oil level and filter in the hydraulic circuit, then reset from the operator panel with the press in manual mode.
P-04 manual · p. 112Alarm reset procedure
3 documents linked to P-04

Machine data, while production is running

Alleantia Core connects machines of any brand and protocol and makes telemetry available in a structured, standardized way. The agents read states, alarms and process parameters and answer with the real numbers from the shop floor.

  • Over 70 protocols and 5,000 machine drivers.
  • States, alarms and parameters, always up to date.
Press P-04In production
Mold temperature182 °C
Cycles per minute14
Hydraulic pressure212 bar
Parts this shift3.912
Alarm E-217 · 14:32

The context that gives numbers meaning

Production orders, downtime reasons and maintenance work explain what happens on the line. With Alleantia Insights and the Connectors to your business systems, the agents know what was being produced, why a line stopped and when the last maintenance was done.

  • Production events linked to machines, lines and shifts.
  • Full history for analysis, comparisons and KPIs.
Line 2 · Shift BToday
  • Order start WO 24-0318
  • Changeover stop · 22 min
  • Maintenance: oil filter replacement
  • Order completed · 4,820 parts
Shift OEE 81%

Shop floor experience stays in the company

Operator notes, fault fixes and best practices go into the knowledge graph. What an experienced technician knows becomes an asset for the whole department, shared according to each user’s permissions.

  • Notes and fixes linked to machines and alarms.
  • Access governed by each user’s permissions.
LBLuca B.Maintenance · Shift ANote
With alarm E-217, check the oil filter first: that's almost always where the problem is. If it's clean, check the temperature probe.
Press P-04Alarm E-217Hydraulic circuit
Added to the knowledge graph
Shared with the team Maintenance
OEE by lineEnd-of-shift reportMicro-stop analysisEnergy use per partAlarm ParetoDowntime by reason
Plant comparisonScrap by batch and recipeSetup times by productMaintenance history by machineMold temperature monitorKPI summary for the shift supervisor

Custom analysis

Create the analyses your factory needs

Describe the analysis you need in plain words: the Coder Agent writes the insight, the Analyst Agent schedules it as a routine and Alleantia Insights runs it on your machine data. Results arrive as reports, notifications or in the dashboard you already use.

  • Ask for it in words, no coding required.
  • Runs on real data, with each user’s permissions.
  • Share it with your team and change it whenever you need.

Everyday support

Less time searching, more time producing

Finding a data point, a manual or an alarm’s history no longer takes hours. The agents know your machines and processes: they answer the shop floor’s questions, reason over the data and prepare the analyses you need. So your team can focus on the work that improves production.

Today · Cutting line
Why did the Cutting line slow down after 2 p.m.?
Analyst Agent
Reasoning
Queries the knowledge graph
Reads the line telemetry
Compares speed and setpoint
Checks the shift's alarms
Since 2:05 p.m., average speed has dropped 18% below setpoint, after three temperature alarms on the cutting unit. Should I save this analysis as a daily routine?
Yes, go ahead
Micro-stops by line
Last 7 weeks
Micro-stops rising
Line 3: +40% over the last two weeks.
Open analysis

Continuous analysis

Decisions made with the data at hand

Many production decisions are still made on gut feeling. The agents run your analyses continuously, surface trends and anomalies and let you explore the data however you like. Every choice you bring to management is backed by verifiable numbers.

The agents

Agents to build and use factory knowledge

Reliable, data-grounded, extensive and fast support, for ad hoc analysis and continuous routines.

Knowledge

Knowledge Agent

Retrieves and manages structured documents and informal knowledge about machines, together with IoT and context data, and creates the AI-ready information layer.

See how it works

Analysis

Analyst Agent

Mediation agent that manages people’s interactions with machine knowledge, Alleantia Insights and LLMs.

See how it works

Vertical expertise Coming soon

Expert Agents

Agents that deliver services to customers, in whole or in part. Each one specializes in a domain: energy, maintenance, performance.

How they work

The agents at work, inside the platform

The librarian and keeper of your industrial knowledge

The Knowledge Agent is the most complex of the group. It analyzes the documents uploaded to the platform and the connected data sources, starting with the telemetry database.

  • Builds and maintains a RAG vector database and a semantic graph.
  • Answers questions about the meaning of concepts relevant to analysis and how they relate.
  • You can query it directly, but it mostly works for the other agents.
Knowledge AgentQuerying the graph…6 linked entities
Which KPIs can we calculate for the Cutting line?
Cutting lineMachine
line_datasheet.pdfDocument
cutting_telemetryTable
part_counterVariable
line_speedVariable
speed_setpointVariable
Setpoint adherenceConcept
Knowledge graph
No document defines official KPIs for the Cutting line. Using the telemetry variables, I can calculate:
Setpoint adherenceline_speed÷speed_setpoint
Parts per hourpart_counter/ hour
Average speed per shiftline_speedper shift
New concept in the graph: Setpoint adherence
Sources: cutting_telemetry · line_datasheet.pdfSample data

Ask questions in natural language and build complex analyses on your data

The Analyst is the agent built for the people who use the platform. Through a familiar chat interface, you can request a one-off analysis or create persistent analysis plans.

  • Analysis plans saved as DAGs of insights, to rerun whenever needed.
  • Calls the Knowledge Agent and the Coder Agent on its own when needed and brings the results together.
  • It’s the interface through which people, and in the future other agents, use the platform.
Analyst AgentBuilding the plan…Plan saved
How did OEE per shift go on line 2 over the last 7 days? Repeat it every Monday.
Analysis plan · DAG
oee_variablesKnowledge Agent
oee_shiftCoder Agent
shift_stopsCoder Agent
monday_reportRoutine
OEE per shift · Line 2Last 7 days
Shift A81%
Shift B76%
Shift C64%
target 75%
MonTueWedThuFriSatSun
Shift C remains 17 points below Shift A. The sharpest drop is on Thursday night, with repeated stops on press P-04.
Routine active: every Monday at 7:00 a.m.Sample data

Governance and security

Factory knowledge, under your control

Choose the model, decide who accesses what and verify where every answer comes from. The agents work on your factory’s knowledge and keep it protected.

Protected know-how

Your documents, procedures and operator experience stay yours. External LLMs never see real names or data values.

Controlled access

Every person and every agent accesses only the data, documents and machines they are authorized for.

You choose the model

Frontier LLMs in the cloud or local LLMs, based on your security and cost requirements. With local models, data never leaves your company.

Answers with sources

Every answer starts from machine data and the knowledge graph and shows where it comes from, so people on the shop floor can verify it.

Transparent insights

The code written by the Coder Agent goes into the Alleantia Insights library: you can read it, edit it and schedule it like any other insight.

Fewer tokens, sustainable costs

The agents optimize model usage and reduce token consumption, even as AI becomes part of everyday shop floor work.

The platform

A complete platform, from machine to AI

Alleantia makes reliable industrial AI possible on the shop floor, with a complete machine-to-AI platform open to customers, integrators and ISVs.

LiveOPC UASiemens S7ModbusFanucOEEEnergyQualityAnalyst AgentWhy is line 2 slowing down?

Let’s talk

What’s the first question
you’d ask your factory?

Book a demo with an Alleantia expert: together we’ll see how the agents can work on your factory’s data and knowledge, with the LLM of your choice.

Book a demo

info@alleantia.com·sales@alleantia.com·+39 050 9911933

A few hours

Installation in your infrastructure, on a couple of virtual machines.

First days

Machines, business systems and documents connected, first agents at work on your data.

About two weeks

A first department with factory knowledge and AI agents at work.

Frequently asked questions

Alleantia Agents in brief

What is Alleantia Agents?

It’s module 3 of Alleantia Industrial Intelligence. It uses the power of LLMs to build and use factory knowledge, with agents for knowledge, data analysis and specialized tasks, supporting the people who work on the shop floor.

Which LLMs can I use with Alleantia Agents?

Alleantia Agents works with any LLM: cloud models or models running locally, chosen according to your security and cost needs.

What's the difference between the Knowledge, Analyst and Coder Agents?

The Knowledge Agent builds and safeguards factory knowledge from documents and data. The Analyst Agent is the interface people use to ask questions and create analyses. The Coder Agent writes and tests the insights the analyses need.

Do I need to know how to code to use the agents?

No. With the Analyst Agent you ask questions in natural language, and the Coder Agent writes and tests the code for the insights you need, directly in the platform.

How are data and intellectual property protected?

The agents access only authorized data and documents, pass only the necessary context to the model and cite the sources of every answer. With local LLMs, data never leaves your company.

Can insights created by the agents be reused?

Yes. Insights and DAGs created by the agents go into the standard Alleantia Insights library and can be edited, scheduled and run like any other insight.