Ferritico
Ferritico is a Swedish cloud platform that predicts steel behaviour with AI. Enter an alloy's chemistry and it returns CCT and TTT diagrams, phase fractions, transformation temperatures and hardness in seconds, before any furnace trial.
What is Ferritico?
Ferritico is a cloud simulation platform for steel, published by Ferritico AB in Stockholm and presented as an AI-powered virtual steel lab. Its purpose is narrow and clearly stated: predict microstructure, hardness and heat-treatment outcomes before anyone runs a physical trial. The company describes the result as a virtual metallurgist working alongside the team.
The core module handles continuous cooling transformation. You enter the chemistry of a steel and the platform returns a detailed CCT diagram in seconds, showing how phases evolve, ferrite, pearlite, bainite and martensite, together with transformation temperatures and hardness for various cooling rates. The practical question it answers is whether a candidate alloy can meet mechanical performance requirements under a given cooling schedule, before a furnace treatment is booked. The web application opens three further modules under a Phase Transformations menu: isothermal transformation (TTT) diagrams, an arbitrary cooling path defined by the user, and batch CCT for running many compositions at once. Tempering effects on microstructure and hardness, hardenability, and surface treatments including carburising and nitriding are announced as additional modules.
What Ferritico claims to bring is hybrid: models trained on decades of real steel experimental data, combined with fundamental metallurgy rather than replacing it. The models are said to be updated continuously with new data and methods, proprietary customer data can be incorporated to refine predictions for specific grades, and a robust API allows integration into existing software and systems. Everything runs in the browser with no installation, teams can tweak parameters, run what-if scenarios and compare alloys in real time, and results can feed FEM simulations downstream.
Around the software sits a consulting arm: strategic road mapping through assessments and workshops to define AI adoption targets and priority processes, custom simulation modules built from customer data up to a digital twin of an alloy system or production environment, and deployment through secure cloud delivery or on-premises APIs connected to FEM software or laboratory databases.
The company is a spin-off from research at the KTH Royal Institute of Technology, registered in Sweden in 2016, and lists a team of six led by chief executive Viswanadh Arigela with co-founders Peter Hedström and Joakim Odqvist. Three subscriptions, Early Technology Adopter, Industrial User and Academic User, are all quoted on request, and a two-week free trial opens when an account is created.
What it does
- Generate a detailed CCT diagram from a steel chemistry in seconds
- Predict ferrite, pearlite, bainite and martensite fractions across cooling rates
- Predict transformation temperatures and hardness for a given cooling schedule
- Produce TTT isothermal transformation diagrams
- Simulate a user-defined arbitrary cooling path
- Run CCT calculations in batch across many compositions
- Feed predicted properties into FEM simulations and related engineering tools
When to use Ferritico / When not to
A quick filter to help you decide if Ferritico is the right fit.
When to use Ferritico
- Steel engineers designing or optimising new alloy grades
- R&D managers who want fewer furnace trials and a shorter time to market
- Heat-treatment and process engineers choosing a cooling schedule
- Materials science researchers and doctoral students, who get a dedicated academic subscription
- Industrial teams needing FEM inputs or a custom model trained on their own experimental data
When not to use Ferritico
- Anyone working outside steel: there is no aluminium, titanium, polymer or composite model
- Buyers who need a published price list, since every plan is quoted on request
- Individuals looking for free self-service access beyond the two-week trial
- Developers expecting public API documentation to integrate on their own
- Engineers who need CAD or structural sizing rather than phase-transformation prediction
How to use Ferritico
A typical end-to-end flow, from setup to results.
- Open ferritico.com and read the Platform section to check that the modules cover your grade and your process
- Either book a CCT demo through the contact form, or use Start Free Trial to reach app.ferritico.com
- Create an account with the single Sign up / Sign in button; authentication runs through Auth0
- A new account starts a two-week free trial, with the days remaining shown on the dashboard
- During the trial, open the CCT module to try the core capability, or the TTT module for isothermal calculations
- Pick a module from the Phase Transformations menu: CCT, Isothermal Transformation (TTT), Arbitrary Cooling or Batch CCT
- Enter the steel chemistry and run the calculation
- Read the predicted phase fractions, transformation temperatures and hardness on the resulting diagram
- Tweak parameters, run what-if scenarios and compare alloys or process conditions with your team in real time
- For a licence beyond the trial, an API integration, a custom model or an on-premises deployment, write to contact@ferritico.com
Pros & Cons
Pros
- Answers in seconds where a furnace trial takes weeks: outcomes simulated in hours instead of months
- Hybrid approach, fundamental metallurgy combined with machine learning trained on decades of real experimental data
- Nothing to install, everything runs in the browser and results are shared in real time across the team
- Unusually broad metallurgical coverage for a single tool: CCT, TTT, arbitrary cooling, batch runs, tempering, hardenability, surface treatments
- A two-week free trial opens on account creation, so the models can be judged on real chemistries
- A dedicated academic subscription makes the platform usable in teaching and research
- European vendor: Swedish law, a detailed GDPR policy with a named supervisory authority, and private-cloud or on-premises options for industry
Cons
- No public pricing at all: the three plans display Contact us and route to a form
- No terms and conditions are published, so nothing covers ownership of results, service levels or termination
- No permanent free plan, only a two-week trial
- The API is a selling point yet has no public documentation
- The sign-in screen restricts account creation to existing customers, so access usually starts with a sales conversation
- No data processing agreement, no security certification, no named subprocessor and no stated hosting country
- No accuracy figures or validity range are published, and the vendor is very small, three employees on the Swedish register
Pricing & Plans
Ferritico publishes no price. All three subscriptions display Contact us and lead to a contact form, so no entry-level amount and no currency can be quoted. There is no permanent free plan: free access is limited to a two-week trial that starts when an account is created on app.ferritico.com. The site invites visitors to contact Ferritico for a pricing consultation, and offers a free consultation on the consulting side of the business.
- single-user access to unlimited CCT and TTT predictions
- price on request
- full access to CCT and TTT predictions plus tempering and hardenability
- with module customisation and API integrations for batch calculations
- price on request
- academic single-user access to unlimited CCT and TTT predictions
- price on request
- Two-week free trial when an account is created on app.ferritico.com
- with the CCT and TTT modules open during the trial
Data, GDPR & hosting
A consolidated view of how Ferritico handles your data.
GDPR overview
Implementation is concrete for a company this size. Ferritico AB is established in Sweden, so the GDPR applies directly and Swedish law governs. The privacy policy runs twelve sections, cites its legal bases by article, Art. 6(1)(a) consent, 6(1)(b) contract performance and 6(1)(f) legitimate interests, and lists every data subject right with its article, from access (Art. 15) to withdrawal of consent (Art. 7(3)), exercised at privacy@ferritico.com. The supervisory authority is named: Integritetsskyddsmyndigheten, Box 8114, 104 20 Stockholm. A four-category cookie banner separates essential, analytics, marketing and functional cookies, and the contact form requires explicit consent. Two gaps stand out: no data protection officer is named, and no data processing agreement is offered. The policy's Last Updated line is generated in the browser and always shows the current date, so no real effective date is published.
Who owns the data?
Ferritico publishes no terms and conditions, so no clause states who owns the chemistry you upload, the diagrams the platform returns, or a model built from your data. The only commitment on the product side is confidentiality: inputs, results and custom models are said to be handled through a secure cloud architecture and kept confidential, and customer data and results are described as isolated, with private-cloud or on-premises options for industrial needs. The privacy policy covers website data only, contact forms, cookies and analytics, and names Ferritico AB, c/o Openlab, Valhallavägen 79 in Stockholm, as data controller, reachable at contact@ferritico.com. Ownership therefore has to be settled contractually.
Reuse rights
Because no terms of use exist, nothing formally grants or restricts reuse of the diagrams and predictions you generate: the site never says whether results may be published, shared with a customer or embedded in a commercial report. In practice Ferritico presents outputs as your own working material, and offers to go further on request, incorporating proprietary experimental data to refine predictions for specific steel grades, or building a model from scratch to create a digital twin of an alloy system or a production environment. That use of customer data is an opt-in engagement rather than a default. Anything about licensing, publication or resale of results has to be written into the contract, since the public pages settle none of it.
Data retention & training
Hosting summary
Ferritico does not say where data is hosted. The privacy policy lists secure data centres and hosting infrastructure among its security measures, alongside SSL/TLS encryption for all transmissions, regular security updates, access restrictions on personal data and regular security audits, but names no country and no region. Subprocessors are given as categories only, cloud hosting providers, email service providers, analytics tools including Google Analytics with consent, and content delivery networks, with no individual supplier named. The jurisdiction, by contrast, is explicit: the controller is established in Sweden, Swedish law governs, exclusive jurisdiction is Stockholm and the supervisory authority is Integritetsskyddsmyndigheten. For industrial customers, private-cloud or on-premises deployment is offered as an alternative to the shared platform. One technical indication rather than a commitment: the web application's public configuration points to a Google Cloud backend in the europe-west4 region, in the Netherlands. The public site sits behind an anycast content delivery network, so the resolved server address says nothing about where anything is stored.
Things to keep in mind
Risks and trade-offs to weigh before adopting Ferritico.
- A prediction is not a trial: the site itself frames the tool as a way to judge an alloy before a furnace treatment, not as a replacement for physical validation
- No accuracy metric and no validity range are published, so there is no way to know where the models begin to drift
- A wrong heat-treatment decision in production has material and safety consequences, and no terms of service define anyone's liability
- Metallurgical judgement can erode if a team stops reasoning and simply trusts the output of the model
- Custom tuning means sending strategic compositions and trial results to the vendor, with no published data processing agreement to frame it
- Ownership of the results is undecided in the absence of terms and conditions, which matters as soon as a diagram is shared with a customer
- Dependence on a three-employee vendor makes continuity of service and access to your custom models a real question
Setup & Integrations
Technical difficulty
Low for the web application: simulations run in the browser with no installation, and a single Sign up / Sign in button handles authentication through Auth0. The real barrier is commercial rather than technical, since the sign-in screen limits account creation to existing customers, so access usually starts with a contact request. Domain skill matters more than IT skill: you need to enter a steel chemistry and read a CCT or TTT diagram, which is engineer territory. API integration, batch calculations, FEM connections and on-premises deployment are engineering projects with no public documentation, handled with the vendor.
Deployment
Supported languages
Behind Ferritico
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What does Ferritico actually predict?
Which modules are available?
Can I try Ferritico for free?
How much does it cost?
Is my data secure with Ferritico?
Is there an API?
Do I have to install anything?
Can Ferritico build a model for my own steel grades?
Who is behind Ferritico?
Is there a mobile app?
Should you pick Ferritico?
Ferritico is a narrow tool that does one hard thing well. Feeding a steel chemistry into a browser and getting back a CCT diagram, phase fractions, transformation temperatures and hardness in seconds is a real change of tempo for an R&D department used to booking furnace time. The metallurgical coverage is coherent rather than opportunistic: CCT, TTT, arbitrary cooling and batch runs are all in the product today, with tempering, hardenability, carburising and nitriding announced alongside. The scientific pedigree is verifiable, a spin-off from KTH research with two professors among the founders, and the vendor is European, which simplifies data protection questions for buyers in the EU.
The reservations are commercial and legal rather than technical. No price is public, no terms and conditions exist, no data processing agreement is offered and the API that features in the sales pitch has no documentation. Nothing on the site states who owns the results, how long platform data is kept, or where it is hosted. The vendor is also small, three employees on the Swedish register, which matters for an industrial buyer thinking about continuity.
The sensible way in is the two-week free trial: create an account, run your own chemistries, and see whether the predictions hold against grades you already know. If they do, the conversation moves quickly to a quote, since every plan goes through the contact form anyway. This is a tool for steel R&D teams, heat-treatment engineers and materials researchers, not for anyone outside that world.
- Choosing a selection results in a full page refresh.
- Opens in a new window.