EUR 2M for Repsense’s Narrative Intelligence and Influence Forecasting

• Lithuanian Repsense builds decision-grade narrative intelligence for defence, governments, and commercial organizations
• The platform combines explainable AI and human analysts to detect, forecast, and interpret influence
• Dual-use approach bridges hostile influence monitoring and reputation or market impact analysis
• The new EUR 2M funding accelerates Repsense’s governance-first expansion toward forecasting, simulation, and decision support

Last December, the Vilnius-based dual-use data analytics startup Repsense closed its Seed round of investment.The platform models how information and narratives spread and influence audiences, for both threat detection and reputation or market impact analysis. The amount of the investment round is EUR 2M and the lead investor is the well known Czech VC firm Tensor Ventures (invested in Anari AI, among others), joined by more proactive CEE investors, such as Seed Starter ČS, BSV Ventures (invested in FirstQFM, among others), and CoInvest Capital (invested in Luna Robotics, among others).

Founders with Complementary Strengths

Repsense was founded in 2021 by co-founders with complementary strengths:

• Mykolas Katkus (CEO) is leading strategy and business development at Repsense while also holding senior roles in PR and consultancy, bringing expertise in narrative analysis and public affairs.
• Alfredas Chmieliauskas (Chief AI Officer) is a former Palantir engineer and data scientist, combining in-depth technical expertise with academic research and AI-driven analytics leadership.
• later joined by Liudvikas Andriulis (CCO) with a background in sales, marketing, and business development, focused on scaling market reach and commercial partnerships.

The name ‘repsense’ stands for ‘reputation sense.’ It self-apparently describes what the platform is meant to do: sense reputation early, quantify it, and track how it shifts.

Media Monitoring and Decision-Grade Intelligence

Mr Andriulis summarizes Repsense’s as moving from media monitoring to decision-grade narrative intelligence through such key milestones as:

• Building a platform that doesn’t just track coverage but detects and structures narratives at scale;
• Validating that the same core engine applies to both commercial reputation/performance communications and defence-grade hostile influence monitoring;
• Proving traction early in a demanding environment, including rapid market adoption in Lithuania and winning major clients quickly;
• Scaling from analytics dashboards toward the next stage: forecasting, simulation, and intervention planning.

Briefly put, what Repsense provides today is an AI-driven information analytics platform that tracks, detects, and predicts how narratives and influence spread across digital and traditional media, helping clients understand emerging threats such as disinformation and brand risks. It also analyzes short-form and algorithmic content often missed by traditional tools and gives clearer insight into narrative impact and audience reactions. This combination supports both security-focused threat detection and commercial reputation or marketing intelligence — for governments, institutions, and businesses alike.

‘Repsense technology shows how quickly DeepTech is penetrating several areas at once. Information is an increasingly powerful asset, and Repsense is a powerful information defense tool, but also a whole new space for advanced marketers,’ at Tensor Ventures’ partner Ondřej Lipold comments.

Trust, Explainability, and Signal over Noise

‘The main challenge has been trust: modern media is noisy, adversarial, and multilingual — so a serious solution must remain explainable, evidence-based, and operationally useful, not a black box. We separate signal from noise using a combination of change + reach + structure,’ Mr Andriulis shares.

Repsense stems from the notion that a narrative matters not because it exists, but because it becomes capable of shaping perception. As such, the platform looks for:

• change points (sudden shifts in volume, framing, or actors),
• reach-weighted exposure (how much real audience it touches),
• network structure (organic spread vs centralised amplification vs coordination),
• persistence and mutation (does it disappear or evolve?)

The underlying model is that narratives become significant when they gain propagation capacity, cross communities, and start producing measurable influence.

Shaping Reputation as a Measurable Signal

‘Reputation today is shaped in real time by narratives in media and social platforms. We built Repsense on the premise that reputation is not a vague concept — it is a measurable signal,’ Mr Andriulis points out.

Here, human analysts are particularly fundamental. AI is used to compress complexity, clustering narratives, detecting anomalies, mapping networks and propagation, and measuring narrative dominance and shifts. It is a co-pilot that proposes signals, while humans validate meaning and relevance. Analysts provide contextual interpretation in culturally sensitive situations, adversarial reasoning (coordination, intent, deception), decision-making under ambiguity.

Language, Culture, and Explainability

Naturally, nuance arises across languages and local media ecosystems. Repsense makes sure to treat language and culture as part of the signal — not as interchangeable text. The platform combines:

• locale-aware modelling and calibration (no one-size-fits-all model),
• entity and concept normalization (aliases, transliterations, local slang),
• confidence scoring and false-positive controls,
• human-in-the-loop validation where stakes are high.

Special attention goes to avoiding the classic NLP failure mode: misreading local nuance and producing false positives.

Overall, Repsense makes sure that every high-stakes finding, such as hostile narratives or influence networks, is explainable at all times. The results are presented:

• what changed, where, and when,
• who initiated or amplified it,
• how it propagated across platforms and communities,
• what examples represent the narrative (posts, headlines, videos),
• confidence levels and alternative hypotheses.

‘We are strict on this: an analyst should be able to explain an alert to leadership on one page, with evidence attached,’ Mr Andriulis states.

Where Defence, Markets, and Governance Converge

Curiously, a lot transfers between defence and commercial use cases:

• narrative detection, clustering, and measurement
• propagation modelling and network analysis
• early warning of emerging shifts
• interpretability and evidence-based outputs

On the other hand, they diverge in:

• objectives (protect society vs protect brand/performance),
• constraints (oversight, escalation risk, attribution sensitivity),
• time horizons (crisis response vs campaign/business cycles).

The core insight is that both domains demand no less than decision-grade narrative intelligence, and mere media clipping won’t do the trick.

‘The information environment has become a frontline for governments, businesses, and institutions alike. We built our platform to give decision-makers the tools to see what’s coming and respond faster,’ Mr Katkus adds.

Mr Andriulis emphasizes that serious institutional users require and serious providers offer:

• traceability and auditability
• evidence and interpretability
• data minimization and privacy-by-design
• governance and risk management

This way, the direction of EU regulation (AI Act, DSA, GDPR) reinforces what Repsense already builds as if outputs were to be scrutinised — because in defence and regulated environments, they will be, the CCO firmly believes.

Long-Term Impact: From Intelligence to Strategic Advantage

Seed Starter ČS actively seeks deep tech companies whose innovations can be applied not only in security and defence, but also across the banking and FinTech sectors.

‘Repsense is building capabilities that will matter far beyond today’s information challenges, and we are excited to support their growth,’ the firm’s CEO Tomáš Milota remarks.

According to Mr Andriulis, Repsense’s new funding accelerates three layers of development over the next 18 months. In the first six months, the focus is on strengthening core narrative intelligence by improving detection and forecasting primitives, enhancing interpretability tooling, and hardening workflows for institutional users.

Over the following six months, the company plans to expand modalities and scalability through richer short-form video and cross-platform analysis, improved multilingual coverage, and deeper integration into client environments.

Next, Repsense aims to deliver a next-stage operational prototype, including DIS-style planning and simulation workflows developed in close collaboration with practitioner teams to ensure real-world applicability.

In summary, Repsense sees itself developing further in two key directions. The primary one is to continue building the platform as dual-use but governance-first — capable systems that remain auditable, responsible, and usable in high-scrutiny environments. Next on is moving beyond monitoring into decision support: forecasting narrative trajectories, comparing intervention options, and measuring effects.

‘This is the point where ‘reputation sense’ evolves from awareness into measurable strategic advantage,’ Mr Andriulis concludes.

Combining explainable AI with human judgment, Repsense treats the information environment not as background noise, but as a measurable, governable domain where real decisions are made and real consequences unfold. Emphasis on evidence, auditability, and cultural nuance sets a higher standard for how narrative intelligence should function in both democratic institutions and competitive markets. In an era where influence spreads faster than regulation or response, Repsense is there to deliver the tools needed to move from reactive awareness to informed, accountable action.