Discover the challenges 2025

Want to scale up your startup into a thriving business?

We unlock the full potential of deep-tech startups by connecting them with investors and corporations across Europe.

Startup-Corporate Collaboration / Challenges

The Startup-Corporate Collaboration begins with a set of challenges presented by corporations, followed by a call for startups developing disruptive technologies that could provide solutions to address these challenges. We evaluate the best proposals and connect both startups and corporations with each other.

The benefits startups will gain:

  • Top-notch mentoring to develop your financial skills and unlock your full fundraising potential.
  • Online support from local mentors and workshops led by some of the big names in the international deep-tech scene
  • Have the chance to pitch your project to top investors, business angels and venture capitalists.
  • Get support from corporates to scale up.
  • Become part of an international network of fellow entrepreneurs, corporates, mentors, and investors with whom you can discuss partnership opportunities.
  • Gain insights about the latest market and innovation trends.

Create new market opportunities and help shape the deep tech startup ecosystem.

Selection Criteria:

Startups interested in applying should meet the following criteria:

  • Focus on Deep Tech: We are looking for startups, university spin-offs, and research teams working in deep tech, developing innovative, technology-driven solutions that push technological boundaries.
  • Strong Core Team: Your team must have expertise in key areas such as marketing, technology, and leadership. A strong team is essential to not only develop your idea but also carry it through to successful implementation.
  • Addressing Real Market Problems: The solution should address a real market need or challenge that can be solved through complex, technology-based innovation. We want to see how your product or service meets actual demand and solves significant problems.
  • Eligibility for Applications: We welcome applications from deep tech startups, university spin-offs, and research teams of scientists who are pushing the boundaries of technological innovation.
  • Geographic Focus: To be eligible, your organization must be based in the European Union, as we are committed to supporting local innovation.
  • Aligning with Market Needs: Your deep tech solution should be closely aligned with specific market challenges and expressed needs, offering a direct and impactful response.

Why should you participate?

  • Potential to Gain a Major Customer: If selected, your solution could be tested by leading companies and institutions, potentially becoming their next innovative solution provider.
  • Valuable Industry Exposure: Your idea will be seen and evaluated by key decision-makers in prominent companies and public institutions, giving you visibility and credibility in the market.
  • Build Strong Industry Connections: Forge relationships with established corporations, opening doors for future collaborations or contracts.
  • Gain References for Your Startup: Even if not chosen as the final solution, you can receive feedback and references that bolster your reputation and future business opportunities.
  • Expert Guidance and Support: Receive advice from atTRACTION experts on planning the testing with challenge providers, legal arrangements, and intellectual property, ensuring a smooth and professional collaboration experience.
  • Enhance Your Problem-Solving Skills: Engage in real-world challenges that sharpen your analytical and creative problem-solving abilities, preparing you for future opportunities.
  • Access to Resources: Benefit from access to tools, resources, and networks that can help you further develop and refine your solution.
  • Potential for Funding Opportunities: Increase your chances of attracting investors who are looking for innovative solutions and talented teams.
  • Contribute to Innovation: Play a key role in driving innovation within your industry by addressing significant challenges faced by leading corporations.

How to Participate:

Step 1: Discover the Challenge

  • Select the challenge you would like to address.
    OUTCOME: Challenge selected by the startup.

Step 2: Apply for the Challenge and Confirm your Application

  • Click the Apply and Solve Challenge button and complete the form until 2025.09.15
    OUTCOME: Registration form completed and solution submitted.

Step 3: Await Approval

  • On 2025.09.30 receive confirmation from the AtTRACTION team that your solution has been approved.
  • Prepare to present your solution to the challenge owner.
    OUTCOME: Challenge confirmed and solution prepared for the challenge provider.

Step 4: Present Your Solution

  • Meet online on Matchmaking Week with the corporate entity that provided the challenge and present your solution.
    OUTCOME: Solution prepared and presented.

Step 5: Confirmation of Selection

  • On 2025.10.13 Receive confirmation from the AtTRACTION team that your solution has been selected as the winner.
    OUTCOME: Best solution selected and announced for the corporation.

Step 6: Finalize Agreements

  • Sign the NDA with the corporation for which you developed a solution and arrange for pilot testing of the solution.
    OUTCOME: NDA signed and pilot testing scheduled.

Step 7: Conduct Pilot Testing

  • Conduct pilot testing of your idea in collaboration with the corporation that presented the challenge.
    OUTCOME: Pilot testing completed.
Apply and Solve Challenge

Discover the challenges

#1 INTEGRATED COMPETENCY VISUALIZATION AND TALENT INSIGHT SYSTEM

Challenge provided by: UAB Reiz Tech​

Competency Visualization Dashboard for Talent Management​

This challenge seeks the development of an intuitive visual dashboard that translates raw organizational data into competency insights for employees and leadership at Reiz Tech.

More information:​

The challenge Reiz Tech is addressing affects the entire organization — from managers preparing annual plans, HR professionals evaluating career paths, and team leads fostering growth, to individual employees tracking their own development journey. Despite its significance, competency tracking often goes unmeasured due to the lack of real-time, structured monitoring tools, becoming a hidden overhead within organizations.

This lack of visibility makes it difficult to identify skill gaps, assess readiness for new roles, or determine where to invest in training. Moreover, some stakeholders remain skeptical about the value of data monitoring in talent development, which underscores the need for compelling, accessible visualizations.

Reiz Tech is developing a customizable and intuitive dashboard solution that translates raw data and complex AI insights into clear, actionable intelligence. These dashboards will visualize individual and team competencies, growth over time, learning progress, and development gaps — supporting strategic decision-making across all levels.

The solution must be simple enough for employees, yet robust enough for executives like CFOs or team leads.

Apply and Solve Challenge

#2 AUTOMATED HERITAGE DIGITIZATION AND ACCESS MANAGEMENT SYSTEM

Challenge provided by: Kauno apskrities viešoji Ąžuolyno biblioteka​

Automated management and access to large amounts of non-digitized documentary heritage​

This challenge seeks the development of an innovative, large-scale digitization and content management system for Ąžuolynas Library.

More information:​

Ąžuolynas Library holds a large amount of non-digitized documentary heritage and extensive information about Kaunas (buildings, people, rivers, etc.) — thousands of boxes and volumes containing manuscripts, old prints, photographs, maps, and other unique documents. Managing, describing, and digitizing these large collections requires significant human and time resources, which slows down their accessibility for researchers and the public.

The sought solution includes:

  • Large-scale and fast digitization: Innovative solutions are needed to efficiently and quickly digitize large quantities of documents in various formats. This could involve next-generation robotic scanning systems that handle documents with minimal contact or advanced image-processing algorithms that optimize scanning quality and speed.
  • Automated content recognition and indexing: Advanced artificial intelligence technologies should be applied for automatic recognition of text (OCR, HTR – handwritten text recognition), images, and other elements in scanned documents. The solution should be capable of extracting and indexing essential information such as dates, personal names, place names, topics, and keywords, even from complex handwriting or old language documents.
  • Data structuring and link creation: After creating digital images of documents and recognizing their content, tools are needed to help structure this data, identify relationships between different documents (e.g., related persons, events, topics), and automatically generate metadata according to international standards.
  • Search and access optimization: An advanced search system is required to allow users to efficiently find information within the digitized collection of documents, texts, photographs, and videos using natural language queries. It should also be able to provide related results even if they are not directly linked by traditional means. A possible feature is a recommendation system that suggests related documents or topics to users.
Apply and Solve Challenge

#3 FEEDBACK ACQUISITION AND ANALYSIS SYSTEM

Challenge provided by: Kauno apskrities viešoji Ąžuolyno biblioteka​

Smart Multifunctional Feedback System at Ąžuolynas Library​

This challenge seeks the development of a smart, multifunctional feedback system for Ąžuolynas Library that enables fast and efficient collection, processing, and analysis of detailed and objective visitor feedback.

More information:​

There are big challenges in quickly and efficiently collecting, receiving, and processing objective and detailed feedback.
A software app is needed that allows the library to gather and analyze visitor insights and feedback about the services provided. This app would use artificial intelligence, natural language processing, and other technologies.

The main features of the app would let visitors give feedback about the library’s services (such as the quality of events, training, educational programs, cleanliness, customer service, etc.) not only in writing but also by voice or through a chatbot. Using AI, the app could detect the visitor’s emotions, verify the accuracy of the feedback, and organize common issues and frequently mentioned topics.

The app would automatically generate reports and analyses that the library can use to improve its services. It would also provide recommendations on how to address the identified problems.

Apply and Solve Challenge

#4 AUTOMATIC SYSTEM FOR RECOGNIZING AND TRACKING ITEM MOVEMENT

Challenge provided by: Kauno apskrities viešoji Ąžuolyno biblioteka​

Automatic recognition and movement tracking of objects (books, newspapers), statistical analysis, and data processing.

This challenge seeks the development of an automated system for object recognition and movement tracking in a library setting, aimed at capturing real-time data on how physical materials—such as books and newspapers—are accessed by visitors.

More information:​

The library collects statistics on which publications people take from the shelves to browse or read on-site (i.e., items that are read but not borrowed). Currently, this data is recorded manually by staff. However, it’s not possible for staff to physically observe where visitors go or which materials they take from specific shelves.

The cameras currently in use do not appear to have the capability to capture or interpret this kind of information.

The main challenge is to determine whether it is possible—and how—to automatically track the use of books and other materials on shelves across two floors of the library. One possible approach could be object recognition (e.g., recognizing books or newspapers).

Each shelf belongs to a specific section. The goal is to gather clear statistics on how often each section is used and which are the most frequently accessed.

Apply and Solve Challenge

#5 AI-POWERED ENHANCEMENT OF IMAGE UNDERSTANDING AND PERSONALIZATION AT SCALE

Challenge provided by: UAB Zedge Lithuania (DataSeeds)​

AI-driven image tagging, semantic segmentation, aesthetic analysis, personalization modeling, and data embedding optimization using large-scale human-voted visual datasets.

This challenge seeks AI-driven solutions to enhance image tagging, semantic understanding, personalization, competition scoring prediction, aesthetic quality enhancement, and embedding optimization using Zedge’s DataSeeds dataset.

More information:​

Despite having millions of human-voted images from themed photo contests, Zedge faces challenges in extracting the full value from this data to improve search, personalization, and content quality.

Solving this challenge would enable more precise and context-aware discovery, boost engagement and monetization through personalized recommendations, support high-quality personalization and discovery through faster and more scalable aesthetic similarity search, and improve visual quality and appeal of wallpapers.

This is Zedge’s problem, impacting their users through search and discovery experiences, and potentially content creators who are aiming for visibility and recognition.

 

Desired solutions:

● Improved image search and retrieval (for example by better auto-tagging. Objective: Enable more precise and context-aware discovery of wallpapers and other media in Zedge’s search and browse experience by automatically tagging images using visual and semantic understanding.

● Image semantic segmentation. Objective: Enhance automated tagging and content understanding by accurately segmenting objects or regions in images.

● Photo Competition Score Prediction Model. Objective: Develop a model that predicts the score an image would receive in a themed photo competition, based on historical data of millions of images and real user voting behavior across contests—each associated with a specific theme.

● User taste modeling for improved personalization. Objective: Predict individual user aesthetic preferences to deliver highly personalized wallpaper recommendations in the Zedge app, boosting engagement and monetization.

● Aesthetic modeling for generative image enhancement. Objective: Use human-labeled voting data to train generative models that improve the visual quality and appeal of wallpapers, through tasks like upscaling, denoising, and style refinement.

● Embedding compression or optimization. Objective: Make Zedge’s large-scale aesthetic similarity search faster, cheaper, and more scalable to support high-quality personalization and discovery.

Apply and Solve Challenge

#6 INTERNATIONAL COMPANY SPECIALIZED IN WASTE MANAGEMENT AND ENVIRONMENTAL PROTECTION

Smart Segregation – Gamifying Waste Sorting & Intelligent Quality Analysis​

This challenge seeks solutions for the problem of low waste-sorting quality by households, it should combine gamification and citizen engagement with intelligent waste-quality analysis to improve sorting accuracy and identify areas needing support.

Low waste-sorting quality reduces recycling rates, increases costs, and exposes municipalities and waste companies to financial penalties. Some materials intended for selective collection still end up in mixed waste. Some households do little or no sorting. Placing recyclables in mixed waste often makes later separation and sale impossible (extracted material is heavily contaminated), preventing effective recycling. Radical improvement of source sorting is a fundamental change that can significantly boost
recovery. Gamification should lead to a reduction in mixed waste and an increase in yellow, blue, green, and brown bag streams.

Failure to meet recycling targets leads to fines, higher processing costs, and reputational loss. Improving sorting quality could significantly reduce costs for municipalities, residents and increase revenue for waste collectors, and recyclers.

International company specialised in waste management and environmental protection is interested in collaboration with a startup, who offers an mobile app with gamification features(points, rankings, rewards, inter-district challenges). The app should be integrated with monitoring and reporting systems for sorting quality. The solution can offer loyalty programs for residents (local discounts, social benefits) and allow interactive educational campaigns using gamified elements. Waste quality monitoring systems in multi-family buildings could allow tracking and identifying contamination patterns at the household or building level (deanonymization), while respecting privacy regulations, and provide municipalities with a real basis to differentiate fees for residents. Considerations: Solution
must combine the above elements to provide actionable data for municipalities to target interventions and improve recycling rates.

Apply and Solve Challenge

#7 WASTE STREAM MORPHOLOGY ANALYSIS

Waste Stream Morphology Analysis

This challenge seeks solution of automated, real-time analysis of waste stream morphology (e.g., black or yellow bags) at waste recovery installation on conveyor belts before further cleaning, providing data to support recycling and process optimization.

Waste management operators cannot rely on morphology analyses, which are manual, costly, and based on small samples. Current methods are unrepresentative and fail to show the true composition of mixed or yellow waste streams. Without these data, it is difficult to plan recycling processes, reduce calorific value for incineration, or improve stabilized waste cleaning. Automated morphology analysis offers accurate data, better decisions, and higher recycling rates.

Companies and municipalities cannot locate sorting errors. Opportunities to recover additional material are lost. Recyclers receive lower-quality materials. Incinerators process waste with too high calorific value. Facilities must account for the entire waste stream uniformly, even if quality varies by municipality.

The potential solution may provide automatic morphology analysis using cameras, sensors, AI, and machine learning. Real-time monitoring of waste composition on conveyor belts is preferred. The solution should allow reporting, visualization, and trend identification (e.g., seasonal changes in black bag composition). Constraints: the solution must be installable in existing sorting facilities, scalable and resilient to harsh conditions (dust, humidity, variable waste composition).

Apply and Solve Challenge

#8 NEW RECYCLING SOLUTIONS OF DIFFICULT WASTE

New recycling solutions of difficult waste

This challenge seeks news solutions to increase recycling rates for hard-to-process, post-sorted fraction from mixed and selectively collected waste, to manage them as secondary raw materials or to create new products that help close the material loop through reuse.

New solutions of processing post-sorted fraction from mixed and selectively collected waste are needed. Currently these fractions are not further recycled, but are instead thermally treated or landfilled. The difficult-to-recycle fractions comprise :

  1. Oversized/bulky fraction generated from mixed waste processing (granulation 80–350 mm),
  2. Stabilized fraction (20–80 mm) after aerobic stabilization of mixed waste – heavily contaminated with plastics, making further processing difficult,
  3. Sorted/bulky fraction from manual sorting of yellow and blue bags, mainly plastics not suitable for recycling plants with energy recovery.

These fractions contain heavily contaminated materials, multi-material waste, and other low-quality residues, which usually end up in landfills or energy recovery plants. This results in higher costs for landfill and waste disposal and fines for missed recycling targets. Secondary materials that could be recovered and sold are lost. The problem affects sorting facilities’ operational efficiency loss with: Longer processing times and less efficient use of sorting facilities, recyclers loosing potential revenues, municipalities, government, which cannot meet recycling targets) and the environment. The possible solutions may include new recycling technologies, new material products or next-generation energy recovery. The solution need to consider the fact that the difficult fractions are heavily contaminated and heterogeneous, containing significant plastic content. Homogenization and processing methods must minimize emissions.

Apply and Solve Challenge