Bioplastics production optimisation
Simulate the process. Improve the decisions.
From fermentation data to production scenarios: machine learning and agentic AI supporting yield, consumption and research.
AI ENGINEERING
Complex problems.AI solutions built for you.
We combine strategy, data and development to improve processes, products and decisions. From machine learning to agentic AI, we select and combine the methods your project needs.
From understanding the problem to a usable solution.
HOW A SOLUTION TAKES SHAPE
The problem shapes the architecture.
Data and results guide the system’s development.
Define what needs to change.
We start with the process, product or decision to improve. We define expected value, constraints and criteria for assessing the project.
Build a usable foundation.
We organise databases, documents and domain knowledge, preparing information for analysis, models and applications.
Choose the intelligence you need.
Machine learning for patterns, optimisation for alternatives, generative AI for text and knowledge, and agents to coordinate actions and tools.
Put models to work.
We develop platforms, interfaces and connections to business systems. APIs and protocols such as MCP make data and capabilities available.
Measure what changes.
We assess performance on representative cases and make results usable by people, with controls appropriate to the context.
HANDS-ON EXPERIENCE
Our consultants’ projects.
AI, quantitative models and systems integration: a selection of projects our consultants have contributed to across different professional contexts. Explore the challenges and the work involved.
27 projects
Industrial data digitalisation
Manufacturing data, finally connected.
CAD drawings, management software and Excel in a shared data foundation: find past jobs, compare costs and put production history to use.
AI assistant for asset data
Ask a question. Explore your assets.
Operational data, energy use and documents become accessible through natural language. A conversational platform developed with a focus on the maritime sector.
Scheduled maintenance platform
Every asset has a history. Every task has its time.
Digital asset records, checklists, history and schedules in a dedicated maintenance platform. The project was tailored to low- and medium-voltage switchgear.
AI estimation of engineering hours
Competitive bids. Engineering effort estimated with care.
A dedicated database and an agentic system to estimate engineering effort for EPC projects, drawing on the MDR and experience from previous projects.
AI technical archive
Your next project starts with what you already know.
An archive you can talk to: find similar projects, retrieve previous solutions and explore historical costs. An approach applied across different settings, from engineering firms to in-house technical departments.
AI for engineering design and simulation
Strengthen R&D by building on engineers’ expertise.
An applied training programme for a highly specialised engineering team: using LLMs, code and AI tools to explore digital models, simulations and new approaches to design.
Humanoid robotics roadmap for manufacturing
Know where to start. Plan how to scale.
A technical and economic assessment for introducing humanoid robotics in a manufacturing company: priority processes, investment scenarios and a path from an initial pilot to wider production use.
Quantitative engine for banking pricing
Estimate the impact of a pricing change before applying it.
Quantitative models for assessing the expected financial impact of repricing accounts and fees. Experience connecting contractual terms, costs and scenarios to support bank decisions.
Integration analysis for an FX pricing engine
Connect pricing complexity with the bank’s systems.
Functional analysis supporting the integration of a complex foreign-exchange trading pricing engine for bank branches in Central and Eastern Europe.
Banking systems adaptation and financial reporting
From regulatory requirements to systems impacts.
Experience in two areas: coordinating MiFID II changes in banking systems and managing the impact of the Short Selling Regulation on short-position reporting.
AI assistant for technical support
Manuals, spare parts and serial numbers. One conversation.
An assistant connects technical documentation, spare-parts availability and work status. Two MCP servers bring knowledge and operational data together for medical-device technical support.
Broadband activation forecasting
Anticipate demand, municipality by municipality.
A predictive model to estimate broadband activations one year ahead, supporting a nationwide service launch across 8,000 municipalities.
PPE and procedure checks with computer vision
Make situations needing review visible.
Deep-learning video analysis to support safety checks in electrical substations: protective equipment and observable steps in operating procedures.
Document data extraction and reconciliation
From documents to the data your systems need.
Neural networks and NLP to extract information from varied delivery notes, transport documents and invoices, then reconcile it across automotive accounting and management systems.
Package and label quality inspection
Spot the error while the package is on the line.
Cameras and computer vision models to detect package non-conformities, check label orientation and support operators’ sorting decisions.
Video creation with AI and media archives
Turn a script into new value for your archive.
Generative AI, automated storyboards and semantic search to connect text and archived content in a personalised video-production workflow.
Cosmetic product and review analysis
Turn the voice of the market into product knowledge.
Ingredients, products and reviews in an AI-accessible data ecosystem for cosmetics marketing and research.
Logistics network optimisation
Design the logistics network before changing it.
Simulation and mathematical optimisation to compare warehouses, flows, deliveries and returns for a multinational fashion company.
Automated property appraisal analysis
From appraisal documents to a property profile.
Generative AI turns information-heavy auction notices and appraisal reports into structured profiles for evaluation.
Genomic analysis platform
Connect genes, conditions and evidence.
The platform integrates clinical, genomic and lifestyle data with literature to explore rare diseases and autoimmune conditions.
AI-powered personalised nutrition plans
Personalise at scale, within real constraints.
Liferapy combines personal profiles, scientific knowledge maps and evolutionary algorithms in a platform that updates plans, recipes and shopping.
NFT and metaverse platform
From a brand idea to a Web3 ecosystem.
A cloud platform connecting NFT collections, community mechanics and e-commerce sites: from template to launch through one coherent experience.
Skills demand intelligence
Understand market demand. Build the right skills.
From LinkedIn job postings to a database accessible through MCP: Job Tower helps identify trends and guide training, service offerings and time-and-materials staffing.
AI assistant for job searches
From conversation to application.
An assistant connected to a recruiting platform guides people towards relevant job opportunities.
Management skills assessment reports
From scores to a development path.
Quantitative assessments become structured reports on strengths, gaps and potential management development paths.
Early identification of at-risk patients
Spot signals that warrant investigation earlier.
Machine learning on hospital data and clinical notes supports the early identification of people at risk of PAH.
A STEP UP IN PROJECT SCOPE
The right technology starts
with the right question.
How can we improve process yield? Where should warehouses be located? What signals are hidden in the data? These challenges call for domain understanding, models and dedicated development.
We design the combination the outcome requires. Strategy, data engineering, mathematical models and software work together. Agentic AI is one option: our scope is the whole problem.
THE CAPABILITIES WE COMBINE
The method follows
the problem.
Each project has its own architecture. We bring together the capabilities needed to build it.
01
Predictive models and optimisation
We develop models to identify patterns, simulate scenarios and find better configurations against objectives and constraints.
WHAT YOU GET
A quantitative basis for evaluating alternatives and guiding decisions.
02
Data and domain knowledge
We connect archives, databases, documents and specialist sources, making information consistent and usable by models and people.
WHAT YOU GET
An information foundation organised around the problem to solve.
03
Generative AI and agents
We design systems that interpret content, generate responses and coordinate tools. We define context, actions and validation points.
WHAT YOU GET
Analysis and interaction capabilities integrated into processes and products.
04
Software and integrations
We build dedicated applications and platforms, connect them to business systems, and prepare deployment, monitoring and ongoing development.
WHAT YOU GET
A solution people can use at work, with dedicated interfaces and integrations.
FROM PROBLEM TO PRODUCTION
A complex project.
Concrete steps.
We define what to build, how to validate it and when to expand it.
Design the system
We assess the objective, data, constraints and integrations, then identify a useful first scope to develop.
Architecture, scope and success criteria.
Validate on your cases
We build a first version and test it on data and situations representative of your work.
A verifiable first result, with clear limits and priorities.
Bring it into production
We complete the software and integrations, configure access and monitoring, and agree on future development.
Release, documentation and an operating plan.
FIND YOUR STARTING POINT
Want to start with
a more focused task?
With AI Automation, we connect and configure the AI tools you already use to carry out defined activities. When you need dedicated software, data architectures and business logic, we build the solution through AI Engineering.
Explore AI AutomationYOUR NEXT PROJECT
Which challenge
do you want to solve?
Tell us about the problem, the systems involved and what you want to achieve. We start there to define feasibility and a useful first result.
The initial discussion clarifies the objective, available data and development scope.
Or email us directlyinfo@icarex.ai