Jobs at Mecka AI

Manage daily operations for Mecka AI’s New York office, including vendors, facilities, supplies, events, visitors, onboarding, and internal processes. Support business development through research, outreach coordination, pipeline tracking, meeting materials, and relationship follow-up. Coordinate across sales, deployments, and engineering while identifying blockers and improving operational efficiency. The role also serves as a trusted operational partner to leadership.

Leads technical data-capture programs across hardware, software, operators, and field locations. Responsibilities include translating dataset requirements into capture plans, coordinating devices and training, validating pilots, investigating data-quality issues, forecasting capacity and delivery, and verifying corrective actions. The role requires analyzing throughput, quality, capacity, and cost while documenting procedures, technical decisions, dependencies, and realistic delivery forecasts.

Own the readiness and reliability of deployed data-capture devices. Diagnose hardware and software issues across cameras, peripherals, Linux systems, storage, networking, and data transfers; prepare and validate devices for deployment; support operators; reproduce defects; create diagnostic scripts and procedures; track recurring failures; and verify repairs, updates, and corrective actions across the device fleet.

Lead end-to-end product development for robotics data infrastructure: turn ambiguous problems into shipped full‑stack products, focus on UX and reliability, define success metrics and instrumentation, and iterate to scale production systems.

Own the end-to-end development and production ramp of tactile gloves, from product requirements and industrial/mechanical design through EVT, DVT, PVT, and volume manufacturing. Lead sensing, firmware, calibration, reliability, supplier, contract manufacturer, production test, quality, cost, and schedule efforts. Coordinate prototype iterations and pilot deployments while ensuring the gloves are comfortable, durable, manufacturable, and produce reliable tactile data at scale.

Own backend services, APIs, SDKs, and operational tools for connected hardware and data-intensive workflows. Design reliable data models and asynchronous systems that handle retries, duplicate events, interrupted transfers, and device disconnections. Build telemetry and diagnostics, maintain production reliability and data integrity, test workflows end to end, and support authentication, versioning, documentation, monitoring, and backward compatibility.

Designs mechanical systems and assemblies for robotics data-capture hardware, taking components from CAD through prototyping, testing, supplier feedback, and field validation. Responsibilities include tolerance and load analysis, sensor alignment, electronics packaging, thermal management, cable routing, serviceability, manufacturing optimization, and reliability improvement. Requires hands-on prototype development, physical testing, engineering drawings, supplier collaboration, and knowledge of materials, fasteners, dimensional inspection, and manufacturing processes.

Designs and automates hardware and software tests to validate recording quality, performance, recovery, and release readiness. Tests long-running sessions and fault conditions, analyzes recorded data for timing errors and corruption, reproduces intermittent failures, isolates root causes, and creates regression tests. Builds physical test fixtures and defines qualification checks across hardware and software configurations.

Develop and debug embedded Linux firmware for camera and sensor recording systems. Responsibilities include board bring-up, reliable data paths, performance and power optimization, recovery from interruptions, field-failure reproduction, and validation across hardware configurations. The role requires strong C or C++ skills, Python or shell scripting, Linux systems knowledge, embedded ARM experience, concurrency and resource profiling expertise, and experience creating reproducible releases and realistic hardware tests.

Leads end-to-end development of hardware systems for robotics and embodied AI, from product definition and architecture through prototyping, validation, manufacturing, and deployment. The role requires hands-on expertise across hardware disciplines, system integration, experimentation, debugging, supplier coordination, and technical leadership.

Designs and validates electronics for robotics data-capture hardware, including schematics, PCB layouts, sensor interfaces, board bring-up, bench debugging, and system validation. Collaborates with firmware, mechanical engineering, and suppliers to improve prototypes, production designs, component qualification, and field performance. Documents measurements and design changes while addressing signal quality, timing, power consumption, and hardware-firmware integration.

Build native Android applications that connect capture hardware, guide data recording, display status, and support recovery from interruptions. Optimize video performance, memory, battery usage, responsiveness, and connectivity across Android devices. Develop diagnostics and automated workflow tests while collaborating with firmware engineers to reproduce and resolve hardware, transfer, and device-interface issues.

Train and evaluate computer vision models for embedded deployment, build efficient C++ inference pipelines, and optimize latency, memory, power, and sustained performance. Integrate models with embedded Linux, test on representative hardware and data, and investigate errors across preprocessing, inference, and system integration. Lead candidates will set architecture and evaluation standards while guiding model and software development.

Develop manufacturing processes, production-readiness and capacity plans for electronic capture hardware. Build and qualify fixtures, calibration equipment, and production tests; analyze yield, cycle-time, and failure data; improve designs for manufacturing; establish repair and rework processes; and guide suppliers through qualification, technical reviews, quality improvements, and production changes.

Build reinforcement learning environments for robotics and embodied AI research. Define tasks, observations, actions, rewards, transitions, termination conditions, and evaluation metrics. Train baseline agents, run controlled experiments, assess generalization and uncertainty, investigate failures, and identify data or environment issues. Collaborate with domain experts and engineers to develop reusable, documented research systems and benchmarks.

Build scalable physics-based simulation environments and world models for robotics and embodied AI. Develop parallel training and evaluation pipelines, calibrate simulations against real-world data, improve sim-to-real transfer, and create learned or hybrid dynamics models for prediction, planning, and control. The role requires strong simulation, robotics learning, Python, and C++ expertise, along with experimental judgment and experience building reproducible research infrastructure.

Lead scientific programs for robotics and embodied AI data generation. Develop operating plans, manage scientific partnerships, launch workflows, coordinate equipment and teams, oversee quality, throughput, schedules, and costs, and improve program performance across locations. The role requires hands-on execution, operational decision-making under uncertainty, vendor and logistics coordination, and practical scientific or laboratory fluency.

Own production online camera calibration, SLAM/VIO systems, and fleet-wide calibration drift monitoring for robotics devices. Develop video-based calibration refinement, maintain calibration databases and version history, coordinate with hardware teams on sensor parameters and synchronization, and build visualization tools for operations and research. The role requires extensive experience with SLAM, VIO, SfM, multi-sensor calibration, optimization, and scalable C++ and Python engineering. It is on-site with periodic travel to a hardware site.

Develop and own classical computer vision systems for 3D hand reconstruction, including multi-view geometry, camera calibration, temporal filtering, inverse kinematics, and physical constraints. Convert model predictions into clean, validated spatial data and maintain scalable delivery pipelines. Build QA metrics, visualizers, dashboards, and failure-analysis tools while partnering with research teams on robotics and embodied AI applications.

Develop generative and physics-based models to infer full-body human motion from egocentric and partial observations. Train motion priors, enforce physical consistency, estimate contacts and ground planes, and deliver MHR/SOMA-X outputs for production robotics pipelines. Define evaluation metrics, commission motion-capture and sensor data, implement research code, and publish results.