The Center for Data Innovation recently spoke with Anupma Kundu, Founder of Calculai, a New Jersey-based company that builds AI-powered tools to help organizations improve decision-making and improve workforce efficiency. Kundu shared insights on how Calculai has applied its AI-driven tools in real-world settings–from smart city infrastructure to healthcare innovation–and emphasized the company’s commitment to building solutions that are intuitive, sustainable, and easy to integrate over time.
David Kertai: What made you want to create Calculai?
Anupam Kundu: I founded Calculai after observing a recurring problem in the AI space: too many organizations were investing in overhyped solutions that promised breakthroughs but failed to deliver lasting value. These efforts often overlooked foundational issues like data quality, organizational readiness, or compliance with regulatory frameworks. Calculai was created to challenge that pattern by bringing clarity, rigor, and realism back to AI integration. At our core, we’re committed to ensuring that AI initiatives meet real-world needs, create measurable impact, and remain viable over time.
Kertai: How do you adapt your AI-driven tools for various industries?
Kundu: We adapt our AI-driven tools by tailoring them to the unique data environments, user needs, and regulatory contexts of each industry. In smart city projects, such as in Cannes, France, we co-developed computer vision systems to analyze real-time crowd behavior and optimize traffic flow during large-scale public events. These tools enhanced safety and responsiveness by tracking how people and objects move through shared environments.
In retail, we leverage behavioral and visual data to personalize customer engagement. AI-driven insights have helped our clients redesign support journeys, improve product discovery, and drive customer satisfaction in ways that feel personal rather than transactional.
For other industries, such as healthcare, we’ve developed AI-driven tools to support clinical decision-making and reduce administrative load. These tools reduce the administration burden on clinicians while maintaining accuracy and interpretability. We also created natural language processing solutions to streamline medical documentation and designed 3D scanning tools using facial recognition to produce custom-fit devices for children with facial deformities.
In every case, we don’t apply a one size fits all model, but instead we build industry-specific AI systems and tools, selecting the right algorithms, training strategies, and feedback mechanisms to fit each client’s workflow, data maturity, and ethical considerations. By aligning the technical architecture with real operational challenges, we ensure that AI becomes a sustainable and trusted tool, not just a technological add-on.
Kertai: How do you tackle both technical and behavioral challenges in AI projects?
Kundu: We approach AI tool implementation as both a technical and human journey. On the technical side, we build systems that are explainable, secure, and easy to integrate into existing workflows. We make sure non-technical users can understand, trust, and monitor how the AI tool operates.
On the behavioral side, we invest time in understanding stakeholder concerns, assessing data governance maturity, and preparing teams for organizational shifts. We co-create solutions with business users, embed trust-building mechanisms, and set up feedback loops to maintain human-in-the-loop oversight. Cultural change is as vital as the code, and we treat both with equal importance.
Kertai: How do you measure the success of your AI initiatives?
Kundu: We measure success through a balanced mix of quantitative and qualitative indicators. On the performance side, we track efficiency gains such as reduced manual work, faster processing, and improved throughput. We also measure accuracy improvements in prediction, classification, or decision-making. Equally important are adoption rates, especially by non-technical teams, and feedback from users and stakeholders. Trust, transparency, and alignment with ethical standards are core to our evaluation process.
Kertai: How do you help organizations achieve long-term success with AI?
Kundu: We assess the long-term adaptability of the solution. Does it evolve as the business grows or as conditions change? We aim to build systems our clients can understand, govern, and improve without becoming dependent on us. Ultimately, success means our clients fully own their AI journey. It’s not just about delivering working models, it’s about empowering organizations to use and evolve them confidently.
