c3

At c3 we're building a spatial intelligence data layer. We're a startup based in the UK, and are building an engineering team to focus on research and development of our proprietary models. 

Ideally we are looking for someone with: 

Multi-session SLAM / map merging / long-term localization / 3D registration experience

Ideal past experience:

Autonomous-driving mapping, warehouse/robot localization, AR persistent maps, LiDAR map differencing, construction progress monitoring, digital twins, or long-term robotics localization.

Requirements

* Strong background in 3D computer vision and geometric perception: point clouds, meshes, depth maps, coordinate transforms, registration, visibility/occlusion reasoning, and spatial change detection.

* Hands-on experience with real RGB-D / LiDAR / depth-sensor data. You should have dealt with noisy depth, incomplete scans, moving objects, partial overlap, and sensor/calibration errors.

* Strong understanding of 3D registration, including ICP variants, robust/global registration, geometric descriptors, RANSAC/TEASER++-style approaches, and failure detection.

* Experience solving multi-session or temporal mapping problems: comparing the same physical environment captured at different times.

* Familiarity with SLAM / visual-inertial odometry. You should understand accumulated drift, scale error, loop closure, and how errors in scan generation propagate downstream.

* Strong Python and PyTorch skills, with experience building production-quality evaluation and inference pipelines.

* Experience with learned feature matching such as LightGlue, SuperGlue, LoFTR, or equivalent is useful, particularly for cross-scan registration.

* Experience with synthetic data generation / simulation for augmentation and controlled failure-case generation.

* Robotics, autonomous systems, mapping, AR/VR, drones, or defence perception experience is highly relevant.

* Fluent English.

* Kyiv-based preferred; hybrid or remote considered.

What you will do

* Build a robust scan0 -> scan1 registration pipeline that aligns mostly-static structure while remaining insensitive to moved furniture and other transient objects.

* Handle partial overlap, occlusion, missing observations, drift, and scale differences between scans.

* Build temporal change detection that distinguishes:

    * object removed / added,

    * object moved,

    * surface or condition changed,

    * area unchanged,

    * area not observable / insufficient evidence.

* Develop visibility and free-space reasoning so that “not seen” is not incorrectly classified as “gone.”

* Build multi-stage registration using geometric and learned features, followed by robust local refinement.

* Define confidence / failure detection for registration and comparison results rather than forcing a result from bad scans.

* Develop condition/defect classification and severity scoring once geometric correspondence is sufficiently reliable.

* Create synthetic perturbations and generated training/evaluation data covering known failure modes.

* Own evaluation against ground truth, including registration success rate, false positive/negative change detection, localization accuracy, and uncertainty calibration.

* Evaluate alternative scene representations, including 3D Gaussian Splatting, NeRFs, TSDFs, occupancy/SDF representations, where they materially improve temporal comparison.

c3 is a research-led startup based in London and Kyiv. We are expanding our team and research and are seeking an RF engineer as part of this expansion. 

Requirements

• Strong background in RF sensing and wireless signal processing: channel state information (CSI), RSSI, multi-carrier/OFDM systems, amplitude/phase extraction, multipath, and the physics of RF propagation.
• Hands-on experience with real RF/wireless hardware data. You should have dealt with noisy channel estimates, phase instability, hardware/carrier frequency offsets, antenna effects, calibration drift, and the gap between clean lab captures and messy real-world signals.
• Strong understanding of RF signal-to-information inversion: extracting physical meaning from raw wireless measurements, and reasoning rigorously about what is and isn't recoverable from a given signal.
• Experience solving temporal sensing problems: detecting sustained change in a physical environment against a learned baseline, and separating genuine events from transient or cyclical confounders.
• Familiarity with the electromagnetic behaviour of materials (attenuation, permittivity, frequency-dependent effects, multi-band sensing) is highly valuable.
• Strong Python and PyTorch skills, with experience building production-quality evaluation and inference pipelines.
• Experience with classical estimation alongside ML — Kalman/particle filtering, Bayesian inference, change-point detection, or equivalent — not only deep learning.
• Experience with signal-processing toolchains (NumPy/SciPy, spectral analysis, filtering) and, ideally, embedded RF platforms (ESP32, SDR, Nordic, or similar).
• Experience with synthetic data generation / physical simulation for augmentation and controlled failure-case generation.
• WiFi sensing, radar, RF localization, IoT sensing, wireless communications, or defence signals experience is highly relevant.
• Fluent English.
• Kyiv-based preferred; hybrid or remote considered.

What you will do

• Build robust detection pipelines that extract a reliable physical signal from cross-link RF data while remaining insensitive to everyday environmental variation.
• Handle noisy channel estimates, phase drift, packet loss, hardware offsets, and differences between devices and deployments.
• Build temporal detection and classification that distinguishes genuine sustained events from transient activity, environment-wide confounders, and cases of insufficient evidence.
• Develop confounder-rejection reasoning so that environment-wide effects are not misclassified as localized events.
• Build multi-band and cross-link inversion using physical models and learned features, followed by robust localization.
• Define confidence / failure detection for detection and localization results rather than forcing a result from bad signals.
• Develop event classification and severity scoring once the underlying signal is sufficiently reliable.
• Create synthetic perturbations and physics-based generated training/evaluation data covering known failure modes.
• Own evaluation against ground truth, including detection rate, false-positive/negative rate per unit-time, localization accuracy, timing error, and uncertainty calibration.
• Evaluate alternative sensing representations and fusion approaches — including multi-band spectral methods, time-of-flight techniques, and integration with external contextual data — where they materially improve fidelity.