Products & Publications
Scholarly publications along with testable products are essential today as they push scientific boundaries and share knowledge across fields. They shape future advancements and support informed decision-making.
Jan 2026
Synthetic Ethical Dialectics Engine (SEDE): An Integrated Approach to AI Governance and Ethical Decision-Making in a Multi-Agent System
The research paper proposes a Synthetic Ethical Dialectics Engine (SEDE), a workflow driven AI governance framework that coordinates ethical debates among AI agents. Each debate stems from a distinct philosophical and technical foundation (utilitarian, deontological, and virtue ethics). The novel workflow integrates these moral philosophies using a reflective equilibrium approach. The proposed engine leverages LlaMa3.2:1b LLM within the n8n development toolkit. The subordinate nodes within the workflow are implemented in JavaScript to enable structured initial analyses, critical conflict analysis, and revised ethical judgments. We demonstrate how moral reasoning benefits a multi-agent system as it brings forth diverse ethical perspectives. SEDE’s outcome is sent in the form of a message to a human moderator for final review via the Telegram app. Thus, the paper discloses the implementation of ethical dialectics in the multi-agent system and further discusses its role for AI safety, policy simulation, and explainable AI ethics.
Dec 2025
Self-Reflective AI Agent for Scientific Hypothesis Generation with Local LLMs
Current AI agents are strong in automation and are popularly regarded as ‘task-followers.’ These agents execute instructions or optimize for a given goal. However, they are weak in generating novel and testable scientific insights. The research paper proposes a self-reflective AI agent that is capable of evaluating its own reasoning, identifying gaps, and generating hypotheses across scientific disciplines. We have developed a prototype ‘Patent Innovation Agent’ with modules for literature exploration (USPTO database), reasoning reflection, hypothesis generation, and novelty evaluation. The prototype generates novel hypotheses in the ‘Intellectual Property’ domain which is not observed in any of the existing state-of-the-art literature. The research introduces a metacognitive AI framework that acts as a co-creator of scientific hypotheses rather than a task executor. The developed architecture thereby augments scientific work of human researchers.
Nov 2025
A Plant-Inspired Alternative to Conventional Memory System Design
Today’s tech and IoT gadgets are power-demanding electronics. These devices need to factor-in the energy consumption parameter to stay relevant in the long-term. One of the key contributors to this changing energy-spectra are the memory systems used by these devices. This includes SRAMs, Flash, or SSDs. The research taps into this complex energy pattern by proposing a unique plant-inspired alternative to conventional memory technologies. Plant-based memory mimics natural processes seen in plants, specifically attributed to memory-like behaviors as observed in the Venus flytrap. We developed a plant-inspired solar powered memory and tested it to identify whether it contributes to reduction in energy consumption while operating in the real world. The results showed that based on the detected environmental conditions the designed prototype activated specific yet different memory responses on Day 1 & Day 2. The proposal thereby showcased a promising futuristic memory framework for the next-gen devices.
Jan 2024
A Sustainable Antenna Design to Enhance Precision Beamforming Capabilities
The adoption of Multi-User Multiple Input Multiple Output (MUMIMO) tech in IEEE 802.11 standards transformed wireless communication, enabling data transmission to multiple users simultaneously. However, traditional antennas struggle to meet MU-MIMO’s precise beamforming needs. Additionally, the environmental impact of non-biodegradable antennas adds complexity. This research introduces an eco-friendly antenna design using biodegradable plant fibers like cotton, hemp, bamboo, and jute. This sustainable antenna not only addresses e-waste concerns but also offers a long-term, eco-conscious solution for precise beamforming. Experimental analysis validates its effectiveness, showcasing the potential to revolutionize wireless tech with minimal environmental impact.
