20 nov
|
Hexagon Multivista
|
Ciudad de México
20 nov
Hexagon Multivista
Ciudad de México
Hexagon Integrated Building Solutions, the leader in reality capture and automated analysis for the construction industry, is looking for a Senior or Staff Computer Vision Engineer to drive, and stay hands on with, the computer vision and machine learning work at the heart of our products. Most of this work is in our Analyze Hub product, which automatically tracks construction progress and detects deviations and clashes from reality-capture data; you will also own the SLAM and spatial-reconstruction work behind GoCapture, our mobile 360 reality-capture product, where smooth, accurate walk paths are everything. Reporting to the Senior Director of Software Engineering and working alongside our Software Architect, you will partner day to day with two senior machine learning and computer vision engineers and help level up engineers across our teams. This is an individual contributor role, not a people-management one, but it carries real ownership: you build, and you also lead CV and ML projects end to end, holding the bar on the quality of results and pushing the work through to commercial release, often with help from people who do not report to you. Above all, we want someone with a commercial, pragmatic mindset who cares about what it takes to make computer vision succeed in a real product, not just in a paper. If you love taking gnarly perception problems from prototype all the way to shipping product, and lifting a team as you go, this is the position for you.
What challenges will you get here:
- Algorithmic Direction: Shape the technical direction of our computer vision and machine learning algorithms, and at the Staff level own the roadmap, balancing research ambition against the need to ship and making the calls on where to invest and where to keep things simple.
- Hands-On Development:
Spend a large share of your time in the code. You will design, implement, and optimize perception algorithms directly, not direct from a distance.
- Commercial Pragmatism: Keep one eye firmly on the product. A typical example is taking a collection of off-the-shelf SLAM algorithms and experimenting with and improving them to hit our real product targets for path quality, runtime performance, and hosting cost, rather than chasing benchmark numbers in isolation.
- Project Leadership: Lead CV and ML projects from first experiment through to commercial release. Turn open-ended research questions into a plan with clear milestones, success metrics, and quality bars agreed up front; keep the work visible; surface risks early; and know when to push on, change course, or stop. Research-heavy work is inherently uncertain, and we need someone who can still make it predictable enough to plan a product around.
- Quality of Results: Own the quality of what we ship. Ensure algorithms are rigorously evaluated, benchmarked against ground truth, and hardened for accuracy, performance, and reliability at scale, so that results hold up on real reality-capture imagery from construction projects across more than 70 markets worldwide, not just on test sets.
- Leading Without Authority: Get results through people who do not report to you. Shipping CV and ML work depends on product, platform, QA, data,
and operations teams across Hexagon, and you will build the relationships, make clear asks, and follow through until the work lands. Driving to Commercialization: Own the last mile. Push work past the prototype through production integration, release, and customer rollout, with a pragmatic eye on what is good enough to ship now and what can wait for the next iteration.
- Analyze Hub Intelligence: Drive the computer vision and ML at the core of Analyze Hub, including automated construction-progress tracking by floor and trade, deviation and clash detection, and analysis of imagery and derived models. This is where most of the work lives.
- SLAM and Spatial Reconstruction: Own the SLAM, pose estimation, and spatial-reconstruction pipelines that power GoCapture, so that 360 walk paths come out smooth and accurate across mobile and 360 hardware.
- Algorithm and Technology Areas: Provide deep technical leadership across the algorithmic domains our products rely on: o Visual and visual-inertial SLAM, structure-from-motion, and multi-view geometry o Deep learning for detection, segmentation, and classification on large-scale image and video data o 3D reconstruction, point-cloud, and photogrammetry techniques o Camera calibration and sensor fusion across 360, DSLR, and mobile devices o UAV and aerial imagery, including photogrammetry and volumetric or earthwork analysis o Making research-grade models run efficiently and reliably in cloud and on-device settings
- Security by Design: Be the advocate our algorithm work needs for security and data protection. Make the secure handling of customer imagery, models, and pipelines a first-class concern across Analyze Hub, GoCaptu
📌 Computer Vision Engineer (Ciudad de México)
🏢 Hexagon Multivista
📍 Ciudad de México