Principal Research Scientist (PhD) (Bengaluru)

Position Overview: We are looking for a visionary Principal Research Scientist to lead core algorithmic breakthroughs at the absolute frontier of artificial intelligence. You will design next-generation foundation architectures, continuous learning systems, and multi-modal alignment protocols aimed at building scalable, personalized, and efficient intelligence. This role is for a premier researcher who wants to move AI beyond static, internet-trained models toward autonomous, adaptive, and resource-efficient systems. Key Responsibilities - Pioneer Next-Gen AI Foundations: Design architectures for emergent reasoning, long-horizon planning, and cognitive synthesis across distributed intelligence nodes - Develop Continuous Persona Alignment: Build high-fidelity, real-time learning frameworks that capture human traits, speech patterns, and decision-making logic without catastrophic forgetting - Human-Centric Conversational AI: Research dialogue systems that understand intent, emotion, and context for natural human-AI interaction - Architect Agentic AI Systems: Formulate robust, secure AI-to-AI communication protocols for autonomous negotiation and workflow execution - Standardized Agent Interfaces: Design MCP and API layers for each model to enable secure tool-use, function calling, and AI-to-AI communication - Agentic Secure RAG: Build privacy-preserving retrieval systems using FHE, MPC, TEE for secure and sovereign memory - Decentralized Data Sovereignty: Integrate cryptographic and federated learning mechanisms that maximize user data ownership - Optimize for Edge & Resource-Constrained Hardware: Compile and compress multi-modal workloads for efficient execution on low-power devices. Focus on SLMs and optimized models - Represent CoRover: Serve as a technical ambassador for CoRover at international tech/AI conferences and research forums. Build academic and industry partnerships - IP Leadership: File full patents and contribute to the companys IP portfolio. Translate research into defensible IP - Research to Impact: Take research from concept to production at scale Required Qualifications - Education: Ph.D. Required in Computer Science, Machine Learning, Artificial Intelligence, or a highly quantitative field from a premier institution - Frontier Systems Focus: Proven research in foundation models, SLMs, multi-modal architectures, or personalized AI - Agentic Frameworks: Deep expertise in lifelong learning, multi-agent coordination, and high-fidelity memory retrieval - Human-Centric AI: Experience in conversational AI, persona modeling, and alignment - Security & Privacy: Research background in privacy-preserving ML, secure RAG, FHE, MPC, or confidential computing - Hands-on Execution: Highly hands-on in PyTorch/TensorFlow/JAX. Capable of training and scaling models - Publication Record: First-author publications in top-tier international research venues - Patents: Should have patents. Preferably granted or filed in AI/ML Preferred Skills - SLMs & Optimization: Expertise in model compression, quantization INT4/INT8, FlashAttention, vLLM, TensorRT-LLM, TensorFlow Lite, XLA, JAX, Pathways, speculative decoding, edge inference - Hardware Accelerators: Experience with NVIDIA GPUs, Google TPUs, and mobile/edge NPUs - Energy Effective AI: Research in reducing compute, memory, and energy for scalable deployment - Conversational Video Pipelines: Real-time digital twins, talking-head generation, or video synthesis - Distributed Systems: DeepSpeed, FSDP, ZeRO, HPC, GPU clusters, Kubernetes - Confidential Computing: Hardware security, differential privacy, decentralized protocols - Research Leadership: Track record of presenting research at international forums Why Join CoRover - AI with purpose and trust: Build technology that serves humanity - Define the future of personalized and efficient AI from day 0 - Work on fundamental research with global-scale impact - Access state-of-the-art research compute - Publish, patent, and present at leading research forums - Generous ESOP for long-term alignment - Mission: Build the ULM - Universal Language Model from Bharat This is not product engineering. This is fundamental research with global scale. .