Top 10 AI Startups to Watch in Slovenia in 2026

Table of Contents

• Introduction: The Terroir of Slovenian AI
• PeakBeats
• Truecall (Ljubljana)
• Centrum Cognitio (Tina)
• TrueCall (Maribor)
• Insur.Cap
• SaleSqueze
• AI Krpan
• Bird Buddy
• Neo Labs AI
• Sunrise Robotics
• Conclusion: The Vineyard, Not the Ranking
• Frequently Asked Questions

PeakBeats

A runner on the Ljubljana trail feels the music shift perfectly with her stride - the beat drops exactly as her cadence increases. This is the promise of PeakBeats, a generative audio platform that treats music as a closed-loop control system. Instead of static playlists, their AI models consume real-time biometric data - heart rate, acceleration, movement - and output a dynamically mixed soundtrack that adapts to the body's state. According to an overview of AI companies in Slovenia by F6S, PeakBeats is carving out a unique niche at the intersection of generative AI and human performance.

The timing is precise. The global wellness market is projected to exceed $1 trillion by 2028, and consumers are hungry for personalization beyond simple step counts. PeakBeats doesn't just track your run - it shapes the emotional and motivational landscape of the activity itself. Their software transforms audio from a passive companion into an active participant, syncing intensity with effort in a way that static playlists never could.

Rather than building a standalone consumer app, PeakBeats is pursuing B2B partnerships with major sports tech platforms. The goal is to become the white-label audio intelligence layer for connected fitness hardware, much like what Slovenia's evolving startup ecosystem is known for - quiet, essential infrastructure. Watch for licensing deals with European sportswear brands and a Series A round in 2027 that could cement Ljubljana as a surprising hub for audio AI.

Truecall (Ljubljana)

Not all data can touch the cloud. For banks, insurers, and law firms operating under the strictest provisions of GDPR and the EU AI Act, sending sensitive documents to external servers is simply not an option. Truecall (Ljubljana) builds context-aware AI agents that run entirely within a client's own infrastructure, processing text and documents without a single external API call. As EU-Startups' directory of Slovenian startups notes, the company specializes in "secure, on-premise AI deployments for privacy-sensitive industries," targeting a market that hyperscalers have struggled to serve.

The differentiation is sharp. While Microsoft Azure AI and Amazon Bedrock offer compliance certifications, they still rely on cloud architectures that many European financial regulators view with suspicion. Truecall's agents are deployed inside the client's own data center, processing sensitive contracts, client correspondence, and internal documents without any data leaving the premises. This is particularly valuable for Central European enterprises that must reconcile EU-level regulation with national banking secrecy laws in Austria, Slovenia, and Croatia.

The opportunity is significant. As European enterprises grow wary of US cloud dominance, Truecall is positioned as a sovereign alternative for the Mitteleuropa corridor. The company leverages Slovenia's strong tradition in language technologies - a legacy of the Jožef Stefan Institute's linguistics research - combined with local data centres that meet the highest security standards. Watch for a partnership with Nova Ljubljanska banka and an expansion into the DACH region, where the appetite for private AI solutions is growing faster than anywhere else on the continent.

Centrum Cognitio (Tina)

The math of burnout is brutal: after a 50-minute therapy session, a clinician typically spends another 15 minutes on documentation. Across a full caseload, that adds up to hours of unpaid administrative work - hours that could be spent actually helping patients. Centrum Cognitio, a Maribor-based healthtech startup, built "Tina" to solve precisely this equation. Supported by EIT Health's HICEE programme, this AI assistant automates session notes and generates personalized treatment recommendations, cutting documentation time to near zero.

The differentiation from generic note-taking AI is critical. Tina is trained specifically on clinical psychology protocols, not general language models. It doesn't just transcribe - it understands therapeutic frameworks and can suggest evidence-based interventions tailored to the patient's presentation. The system is designed as a co-pilot, not a replacement: the therapist stays fully in control while the machine handles the paperwork. According to F6S's overview of Slovenian AI companies, Centrum Cognitio is already deployed in dozens of pilot clinics across Slovenia and Croatia, focusing on alleviating the administrative burden that drives so many therapists out of the profession.

The timing is acute. Mental health demand is exploding across Europe, yet therapist burnout has reached crisis levels - too many patients, too few clinicians, too much paperwork. Centrum Cognitio's value proposition is unusually clear: help the therapists who are drowning in documentation to focus on what they trained for. Watch for a Series A round led by healthtech venture capital firms and expansion into German-speaking markets, where therapy regulation is famously strict but AI-assistive tools are just beginning to gain regulatory acceptance.

TrueCall (Maribor)

A video call is a strange stage: we speak with conviction while our faces tell a different story. TrueCall, based in Maribor, built its technology on this gap. The platform analyses micro-expressions, facial movements, and gestures in real time during video conferences, providing emotional transparency that words alone cannot convey. Founded by Oleksandr Iurkin, Viacheslav, and Dmytro, the startup raised approximately $125,000 in initial rounds to develop a system that is deliberately non-invasive - it works with standard webcams and requires no wearables or special hardware.

TrueCall does not claim to be a lie detector. Instead, it flags discrepancies between what someone says and how their face moves, offering a layer of insight for corporate negotiation training and fraud detection. The algorithm is trained on a proprietary dataset of cross-cultural micro-expressions, making it more robust than generic affect recognition APIs that fail when confronted with different emotional norms across regions. As F6S's directory of Slovenian AI companies notes, TrueCall is already serving B2B clients looking for more transparent communication in high-stakes negotiations.

The market for conversational intelligence is growing rapidly, particularly in sales training and remote hiring where understanding subtext is critical. TrueCall is still very early, but its potential lies in distribution partnerships. If it can integrate with a platform like Zoom or Microsoft Teams, it could scale overnight. Alternatively, the technology represents an attractive acquisition target for a larger HR analytics or security firm looking to add an emotional dimension to their suite. The key question is whether the startup can move fast enough to capture a market where giants are already circling.

Insur.Cap

Insurance underwriting has long been a game of forms and formulas - text-based questionnaires that capture what a customer reports, not what is actually visible. Insur.Cap, founded by Aleš and Aleš in Ljubljana, flips this paradigm entirely. Their AI analyses images - a photo of a roof, a factory floor, a warehouse - and generates underwriting insights from visual data alone, fusing image features with textual descriptions to predict the likelihood of a claim before the policy is ever written. According to F6S's overview of Slovenian AI companies, the startup is supported by the Bright Labs Incubator (2024) and is developing proactively risk-predicting chatbot interfaces for B2B clients.

The differentiation from the insurtech crowd is striking. Most startups in this space focus on pricing algorithms or chatbot-based claims handling - the easy layers. Insur.Cap tackles the hardest part of underwriting: physical risk assessment. The same image that an adjuster would inspect for damage, Insur.Cap inspects for risk before a policy is issued. A crumbling chimney, an overhanging tree limb, a poorly sealed warehouse roof - these visual signals are invisible to traditional underwriting forms but immediately apparent to Insur.Cap's computer vision models.

The team is now exploring a chatbot that lets customers upload images and receives proactive risk predictions in return - essentially turning the smartphone camera into a pre-underwriting tool. If Insur.Cap can integrate with EU-wide insurance platforms like Guidewire or Duck Creek, it could become the default computer vision layer for commercial property insurance across Central Europe. The B2B pilots with regional insurers are early but promising, and the technology addresses a pain point that every carrier feels: the gap between what a form captures and what a photograph reveals.

SaleSqueze

For a salesperson juggling dozens of custom product configurations, every quote has been a slow and painful exercise. Complex, customizable products - industrial machinery, modular kitchens, solar installations - traditionally required back-and-forth emails, manual calculations, and days of waiting. SaleSqueze, based in Ljubljana, replaces that entire workflow with a 3D Visual CPQ (Configure, Price, Quote) platform. Their generative AI turns a handful of parameters into a photorealistic product visualization, with instant pricing and quotes generated on the spot. According to BeBeez International's profile of Slovenia's evolving startup scene, SaleSqueze is supported by Underline Ventures and was named among the 2026 "top 10 promising startups" in the country.

The differentiation from traditional CPQ tools is stark. Most are text-heavy, rigid, and difficult for sales teams to adopt. SaleSqueze makes the process visual and intuitive, letting a salesperson show a client exactly what they’ll get before any documents are signed. Quote generation collapses from days to hours. This is especially valuable for the 60% of European manufacturers producing high-mix, low-volume goods - factories that need to switch from circuit boards to sensors between shifts and cannot afford a multi-week quoting process for each new product run.

The European manufacturing tech stack is ripe for AI-powered sales tools, and SaleSqueze competes with giants like Salesforce while maintaining a niche focus on physical products. The company is pursuing integration with ERP systems such as SAP and Microsoft Dynamics, and planning a push into the DACH region - Germany, Austria, and Switzerland - where precision manufacturing is the backbone of the economy. As noted in EU-Startups' coverage of Slovenia's most promising startups, the combination of visual configuration and instant pricing is exactly what complex manufacturing has been waiting for.

AI Krpan

Slovenian law lives in a dense thicket: national statutes, EU directives dating back decades, and the evolving case law of the Court of Justice of the European Union. For a lawyer trying to prepare an appellate brief, the research alone can consume days. AI Krpan, named after the folk hero who defended Slovenia against the Ottomans, is a legal AI companion built specifically for this complexity. It analyses appellate court decisions, drafts summaries, and flags inconsistencies in complex legal documents across the full corpus of Slovenian and EU legislation. Based in Maribor, the company focuses on custom models for enterprise clients rather than a one-size-fits-all product.

The differentiation from general-purpose LLMs is fundamental. While ChatGPT or Claude can summarize a contract, they cannot reliably answer the question every lawyer asks: "Can I cite this in court?" AI Krpan is trained on the specific corpus of Slovenian law, EU directives, and CJEU case law, providing legally sound analysis that meets the evidentiary standards of civil-law jurisdictions. As Slovenia's Maribor startup ecosystem matures, this kind of domain-specific vertical AI is exactly where the country's deep linguistic and legal expertise can create an enduring competitive advantage.

Legal AI is a hot category globally, but most players focus on US or UK common law - a fundamentally different system from the civil-law traditions of continental Europe. AI Krpan has a natural moat in multi-language European markets where both national and EU law apply simultaneously. Already deployed in Slovenian law firms, the company is preparing for expansion into Germany, Austria, and Croatia. Watch for a Series A led by a legal-tech venture capital firm and a partnership with a major law firm in Vienna or Munich - the kind of quiet, foundational move that makes an AI system indispensable before anyone notices it has arrived.

Bird Buddy

The simplest joys often produce the most unexpected AI. Bird Buddy, born in Ljubljana, is a smart bird feeder that uses an onboard camera and computer vision to identify species, capture photos and videos, and send notifications to your phone. What looks like a garden accessory is actually a sophisticated edge AI device: it processes species recognition locally, without requiring a cloud connection for basic identification, enabling low-power operation in remote gardens where WiFi is weak or absent. The experience is gamified - users collect digital cards of species they've spotted, share highlight clips with friends, and track seasonal migrations across their own backyard.

The secret sauce is the data. Bird Buddy has amassed a proprietary dataset of millions of labelled bird photos contributed by its global user base, enabling species recognition accuracy that consistently beats generic image classifiers. According to F6S's ranking of Slovenian AI companies, the startup raised significant capital from 10X Capital and others following a record-breaking Kickstarter campaign that exploded into retail expansion across US and EU markets. The hardware-razor-blades business model - sell the feeder, then offer subscription features for advanced analytics - has proven remarkably effective for building a recurring revenue base.

The real prize, however, is not the hardware revenue. It is the user-generated dataset itself, which has become one of the largest labelled image collections of wild birds anywhere in the world. That data has value far beyond consumer entertainment: it can be licensed to ornithology institutes for migration research, or used to train agricultural models that help farmers distinguish pest species from beneficial birds. As Seedtable's overview of Ljubljana's best startups notes, Bird Buddy's combination of consumer appeal and data moat makes it one of the most intriguing companies to emerge from the Slovenian ecosystem. An IPO by 2028 is plausible.

Neo Labs AI

Enterprise customer support is drowning in repetition. Most chatbots follow rigid scripts that frustrate callers and escalate everything to human agents. Neo Labs AI, headquartered in Ljubljana, built something fundamentally different: intelligent AI agents that handle over 100 concurrent conversations across multiple languages, routing only the genuinely complex queries to humans. Their platform covers inbound customer support, outbound sales outreach, and internal help desks, claiming a 70% reduction in operational costs and a 95% customer satisfaction rate for early deployers. According to PRUnderground's profile of Neo Labs, the company is already positioned as "Slovenia's next potential AI unicorn."

The differentiation is architectural. Most conversational AI platforms are chatbots with a fixed script - they recognize keywords but lack context. Neo Labs agents are proactive and context-aware, continuously learning from past interactions to improve accuracy. Their NLP models are fine-tuned on industry-specific corpora - finance, telecom, healthcare - without requiring expensive custom training per client. This means a bank deploying the platform sees it already understanding financial terminology, while a telecom provider gets a model fluent in network issues and billing complaints from day one.

The risk is scale: global giants like Zendesk, Intercom, and Salesforce are pouring billions into similar capabilities. But Neo Labs moves faster and offers deeper customization for Central European enterprises that need nuanced language support and compliance with local data regulations. Watch for a Series C round and a potential IPO within three years, with a major partnership with a regional telecom operator as the likely catalyst for exponential growth.

Sunrise Robotics

The European factory floor faces a contradiction: demand for customization is surging, but industrial robots were built for repetition. Traditional automation excels at doing one thing millions of times - attaching the same screw to the same chassis, every second of every shift. But 60% of European manufacturers need flexibility: small batches, frequent changeovers, products that differ from one hour to the next. Sunrise Robotics, founded in 2023 by Tomaž Štolfa, Joe Perrott, and Marko Thaler in Ljubljana, built modular two-armed robotic workstations specifically for this world. Each cell learns its task entirely in simulation - a digital twin - before it ever touches a physical part. According to Seedtable's overview of best startups in Ljubljana, the company secured €7.2 million from Plural Platform, Seedcamp, and Tapestry VC to bring this vision to scale.

The differentiation from traditional industrial robotics is fundamental. A typical robot arm requires weeks of on-site programming and dedicated tooling for each new task. Sunrise's cells switch from assembling circuit boards to packaging sensors in the same shift, with no hardware changes. The system learns by demonstration and reinforcement learning, not by coding. Deployment takes under 10 weeks from order to production - compared to six months or more for conventional automation. As EU-Startups highlighted in their roundup of Slovenia's promising startups, Sunrise Robotics is solving the most acute pain point in European manufacturing: labour shortages combined with the need for flexible, reconfigurable automation.

The long-term vision is a robot-as-a-service model where factories pay per hour of uptime rather than purchasing expensive equipment outright. This lowers the barrier for small and medium manufacturers - the backbone of the Central European industrial economy - to adopt AI-driven automation. With €7.2 million in seed funding and a clear product-market fit validated by factory deployments, the company has positioned itself as the most physically tangible AI company in Slovenia's ecosystem. A Series A in the €15-20 million range is likely in 2027, with the potential to become the next Comtrade or Outfit7 in scale.

Conclusion: The Vineyard, Not the Ranking

Looking at these ten companies - a legal AI, a bird feeder, an emotion detector, a therapist's assistant, an industrial robot - it is tempting to ask which is "best." But the question reveals more about the asker than the answer. Slovenia's AI ecosystem in 2026 is not a competition. It is a vineyard. Each startup grows in a different soil: Sunrise in the bedrock of Central European manufacturing, Bird Buddy in the consumer garden of global nature lovers, Neo Labs in the rich loam of enterprise communication. The Vega supercomputer at the Jožef Stefan Institute and the new Slovenian Artificial Intelligence Factory (SLAIF) under the EuroHPC Joint Undertaking provide the climate for all of them to thrive.

The real transformation, as one industry observer noted, is happening "quietly in the background" - by the time these startups become famous, they have already become essential. As The Region's analysis of Slovenia's AI strategy points out, the country's approach - slower, inclusive, regulation-heavy - stands in stark contrast to American speed or Chinese scale. Whether that model can compete globally may not be the right question either. Slovenia is not trying to win the race to general intelligence. It is building specialized intelligence for specific soils: a factory in Maribor, a law firm in Ljubljana, a garden in Bavaria.

Next time you read a Top 10 list - including this one - do not ask who is first. Ask: what kind of world would need all of these? The answer is what makes Slovenia's AI scene worth watching. A world where a single country of 2.1 million can grow robotics, legal AI, generative audio, emotion recognition, and mental health tools under one sky is not a world of winners and losers. It is a world of terroir. And the vintage is just beginning to show its character.

Frequently Asked Questions