Tatras Data, founded in 2012, operates as an enterprise AI and data science firm that converts advanced AI concepts into production-grade systems delivering measurable business impact. It specializes in generative AI, predictive analytics, agentic AI, and bespoke machine learning solutions; consequently, it enables organizations to transition decisively from experimentation to scalable deployment. Moreover, through a production-first methodology, Tatras Data embeds directly within client teams, thereby accelerating the design, build, and operationalization of AI solutions, often within weeks. As a result, it serves global enterprises across industries with precision, combining deep technical expertise with sharp business alignment to drive efficiency, unlock innovation, and secure sustained competitive advantage through applied artificial intelligence.
In an exclusive interaction with The Interview World at the 33rd Convergence India Expo, Kiritee Karunya, Principal Data Scientist at Tatras Data, outlines the core capabilities of Koios and the challenges it addresses. He further evaluates impact of the product on recruitment, explains how the product drives operational efficiency by leveraging AI, and details the innovations he envisions over the next five years. The following are the key insights from this discussion.
Q: Can you detail the core capabilities of Koios and explain the specific business or technical challenges it is designed to solve?
A: Koios is a recruitment-focused platform designed for the post–generative AI landscape. As candidates increasingly exploit GenAI tools to game interviews and assessments, the platform addresses this integrity gap head-on. It delivers an end-to-end solution: it generates job descriptions, curates role-specific questions, invites candidates, and records their responses. Simultaneously, it captures voice and video data, enabling a multidimensional evaluation of each candidate.
Crucially, Koios analyses behavioural and cognitive signals in real time. For instance, it detects abrupt response patterns, irregular speech flow, and inconsistent eye movements, common indicators of prompted or assisted answers. Consequently, recruiters gain deeper visibility into how candidates think, respond, and articulate under scrutiny. Beyond correctness, the platform evaluates response quality, coherence, and depth. It also dynamically adjusts question complexity, thereby creating an adaptive assessment that reflects a candidate’s true capability.
Moreover, Koios builds a holistic candidate profile by synthesizing insights across modalities, text, voice, and visual cues. It then compiles a comprehensive, standardized report, which significantly reduces reliance on manual intervention. At the same time, the platform enforces rigor in evaluation through structured grading frameworks. Instead of subjective judgments that vary across interviewers, Koios applies predefined rubrics aligned to specific competencies. As a result, it ensures consistency, fairness, and precision in candidate assessment.
Ultimately, Koios centralizes control within a unified system. It empowers recruiters with transparent standards while mitigating the unintended consequences often associated with agentic AI systems.
Q: Could you elaborate on the impact of your product on the recruitment process?
A: Across sectors, not just HR, the impact is significant. Traditionally, recruiters spend weeks sourcing and screening candidates; even then, they lack certainty about selecting the best fit. In contrast, AI-driven systems compress this timeline dramatically. Within minutes, they retrieve and evaluate candidates, isolate strengths and weaknesses, and present a clear comparative assessment. As a result, recruiters can make faster, more informed decisions about advancing candidates to subsequent rounds.
Moreover, these platforms do more than accelerate screening. They enable precise, data-driven shortlisting by systematically highlighting differentiators across applicants. Drawing on experience across multiple HR tech deployments, such systems consistently achieve up to 60% evaluation accuracy within the first minute of analysis. In parallel, they integrate seamlessly into a comprehensive applicant tracking system (ATS), thereby streamlining the entire recruitment workflow.
Consequently, recruiters gain both speed and clarity. They reduce manual effort, improve decision quality, and enhance overall hiring efficiency. In practical terms, this represents a decisive and highly positive shift for talent acquisition.
Q: How does your product enhance operational efficiency for recruiting organizations?
A: We work with recruiting firms that specialize in hiring for VP- and CXO-level roles. At this level, raw technical skills alone do not suffice; instead, organizations must evaluate leadership, judgment, and strategic impact. However, leadership rarely translates directly onto paper. Therefore, we designed the platform to assess candidates beyond conventional resumes.
Specifically, the system defines structured competency frameworks. It distinguishes between baseline and ideal capabilities, and then maps candidates against these benchmarks. Subsequently, it analyses publicly available data, such as professional profiles, company websites, and platforms like LinkedIn, to build a more complete view of each individual’s track record. Based on this evidence, the platform scores and ranks candidates, thereby surfacing those who demonstrate stronger leadership signals and executive potential.
As a result, recruiters gain a more objective and data-backed evaluation of senior talent. At the same time, the platform significantly reduces manual effort, accelerates decision-making, and improves the precision of executive hiring.
Q: What strategic innovations or capabilities are you planning to build over the next five years?
A: We aim to address a critical limitation in AI systems: hallucination. Accordingly, we are engineering the platform to minimize spurious or ungrounded outputs. At the same time, we are optimizing system architecture and workflows to accelerate execution. Consequently, the platform will deliver faster, more reliable, and contextually accurate results.
