Lichingngamba Tensubam

• Conversational Agent Development: Spearheading the design and development of an intelligent conversational agent for OVNG, capable of classifying user query intentions and responding with real-time data. The agent facilitates seamless communication across the OVNG Network Management platform using REST API-based interactions.

• Dynamic Tool Selection via Custom MCP Server: Engineered custom MCP (Model Control Protocol) tools that integrate with the OVNG ecosystem, enabling dynamic, query-driven selection of tools in real-time to support user intent.

• AI/ML Integration Expertise: Collaborating within a high-performance, two-member team to craft robust, ML-powered software solutions tailored for operational efficiency and intelligent automation.

• Machine Learning & Software Development: Driving the innovation and enhancement of AI-powered features through end-to-end ML pipelines and contributing to the continuous evolution of product capabilities.

• Engaged in AI/ML initiatives, collaborating with cross-functional teams to develop machine learning applications.

• Extracting valuable insights from highly complex raw realtime telemetry data sets to detect Anomalies and Root cause of Anomalies.

• Successfully design PoC for Anomaly Detection and Root cause analysis in telemetric datas in real time.

• Research on Transformer Models for Biomedical NER: Explored and evaluated transformer-based approaches to enhance entity identification and classification in unstructured biomedical text.

• Fine-Tuning LLMs for Medical NER: Optimized BioMegatron using the NeMo framework to improve the recognition and classification of medical terms in unstructured documents.

• Text-to-Speech (TTS) Development: Built a TTS model leveraging SSML for enhanced speech synthesis.

• Containerized Speech Synthesis: Developed a Dockerized solution to generate speech from user-provided text using SSML.

• Model Integration: Implemented Coqui-TTS for high-quality text-to-speech conversion.