Jobs at Hyphen Connect Limited

Optimize and compress large language and vision models for on-device inference. Build model distillation and hardware-specific compilation pipelines, and benchmark performance across various NPU/GPU architectures to enable efficient edge deployment.

Design and architect multi-agent orchestration frameworks, build robust tool-calling pipelines and reasoning loops, and optimize latency and token usage for real-time agentic workflows. Apply expertise in multi-agent frameworks and agentic RAG to enable autonomous tool invocation and scalable workflow automation.

Design and orchestrate large-scale LLM pre-training across 1,000+ GPUs using PyTorch, DeepSpeed, or Megatron-LM. Optimize InfiniBand/RDMA networking and memory to avoid OOM, automate checkpointing and failure recovery for month-long runs, and manage SLURM or Kubernetes GPU clusters. Implement systems-level improvements using C++, CUDA, and Python.

Conduct research developing foundation models for robotics, align text-based reasoning with low-level control policies, design experiments evaluating zero-shot generalization, integrate models into robotic applications with interdisciplinary teams, and present findings at conferences.

Lead product strategy and roadmap to convert advanced AI and robotics research (e.g., VLA models, agentic RAG) into customer solutions. Define safety, usability, and performance metrics, engage early enterprise adopters for feedback, and coordinate cross-functional teams to deliver hardware and deep-tech software products.

Design MAPF algorithms, build low-latency Robot-to-Cloud and Robot-to-Robot communication, and develop cloud dashboards for fleet monitoring and teleoperation handoffs to optimize robotic fleet efficiency.

Design and implement functional safety systems for robotic products ensuring compliance with ISO 13849 and ISO 3691-4. Implement redundant safety-rated PLCs and collision-avoidance logic. Perform rigorous FMEA on hardware and software and support systems engineering and testing.

Design and implement Visual-Inertial SLAM pipelines, develop dynamic obstacle avoidance and predictive path-planning, and optimize navigation algorithms for low-compute edge deployment on robotic platforms.

Build and operate MLOps and agentic infrastructure: manage model registries, continuous training loops, and A/B testing; deploy agent microservices on Kubernetes; and create observability dashboards tracking token usage, latency, and agent reasoning.

Develop intent-recognition models from gaze, gesture, and voice data; implement socially aware navigation that respects personal space; and design visual and auditory feedback for robot user interfaces. Run user studies and rapidly prototype AI-driven voice and visual interfaces.

Design and implement advanced control algorithms for bipedal locomotion and manipulation. Train RL policies in simulation and deploy on hardware, perform sim-to-real transfer, and analyze gait and kinematics telemetry to improve performance.

Design and implement high-fidelity spatial mapping and real-time object recognition pipelines; integrate vision outputs with robotic manipulation and navigation systems; and train and fine-tune Vision-Language-Action models to improve robotic autonomy and perception.

Design and integrate vision and audio models into agent reasoning loops, optimize low-latency voice-to-voice streaming, and architect multimodal RAG systems for retrieving insights from videos and PDFs.

Perform adversarial testing on language models and multimodal agents, implement guardrails and real-time filters for autonomous tools, and develop constitutional AI principles and RLHF alignment pipelines to ensure safe AI deployment.

Design, test, and optimize electromechanical systems and motor designs. Develop and tune motor control firmware (FOC) for BLDC motors, specify harmonic drives and gearboxes, and collaborate with mechanical engineers on joint and thermal management.

Process high-resolution tactile and force-torque data, implement slip-detection and compliant control, train multi-fingered multi-finger grasping policies for unknown objects, integrate tactile sensing into robotic platforms, and conduct experiments to validate tactile perception and dexterous manipulation.

Define and execute the product roadmap for AI-driven on-chain monitoring tools to detect blockchain threats. Collaborate with engineering, research, and stakeholders, analyze market and customer feedback, prioritize features, ensure timely delivery, and maintain transparent communication across teams.

Drive institutional liquidity and TVL by sourcing, pitching, and onboarding market makers, CEX VIPs, prime brokers, and large allocators. Build strategic partnerships, structure and close liquidity incentives, and collaborate with product/tokenomics to design institutional yield strategies tied to TVL KPIs.

Lead media strategy and narrative for a Layer 1 blockchain project. Build reporter relationships, run PR campaigns, draft press releases and thought leadership, manage reputation and rapid-response messaging, and secure high-profile speaking and media opportunities.

Lead full-cycle accounting for fiat and BTC transactions, manage global crypto-to-fiat payroll, oversee liquid assets and on-chain operations, implement crypto-native accounting software for reconciliation and reporting, and ensure compliance with international tax and reporting requirements for a Bitcoin Layer 2 project.