Hire this AI Trainer
Sign in or create an account to invite AI Trainers to your job.
Montreal, Canada
AWS SageMaker
CVAT
Encord
Kili Technology
Labelbox
Label Studio
SuperviselyNo subject matter listed
Document
Image
TextXGO Data Services (XDS) is the AI-services arm of XGO Logistics, a Montreal-based fleet-for-hire carrier that has moved >90 % refrigerated food loads across North America since 2017. Leveraging the same operational discipline that keeps our trucks on schedule, XDS delivers cost-efficient, on-demand data labeling for computer vision, NLP and document-AI projects. We maintain bilingual project managers in Canada for client communication and a vetted workforce in India for scalable production, giving you round-the-clock throughput and up to 60 % cost savings compared with traditional bundled-platform vendors . Every task goes through a three-tier QC pipeline (worker → QA lead → project manager) and we plug directly into the tool you already use—CVAT, Labelbox, Encord, or an in-house interface—so you keep full data control . Our mission is simple: “Move data with the same reliability we move freight—safely, accurately, and on time.”
Physical security. Our India production hub is a secured facility with 24 × 7 CCTV, biometric gate control, and visitor logs retained 1 year. Local ISP lines are dedicated and separated from public Wi-Fi, mirroring best-practice BPO standards. Endpoint controls. Annotators use company-issued laptops with full-disk encryption and USB lockdown; admin rights are prohibited by policy. Workstations auto-lock after 3 min of idle. Network safeguards. All staff connect through an always-on VPN + MFA; outbound traffic is filtered by Geo-IP rules and Layer-7 firewall. Tool isolation. Projects run inside client-provided platforms (Labelbox, CVAT, Encord) with role-based permissions and audit trails, eliminating local data storage. Confidentiality & training. Every employee signs an NDA plus a data-privacy rider in line with GDPR Art. 28. Quarterly refreshers cover privacy, phishing awareness, and secure-coding basics for scripts we occasionally build to speed annotation.
Transcribed 3 000 CORD receipts; reached 95 % field-level accuracy in Labelbox with hotkey templates.
Exported daily trip-detail CSVs from Motive; tagged idle >10 min, harsh brake, geofence arrival for 25 tractors. Pipeline captures ~1 200 events/day and synchronises with dispatch base.
Annotated 5 000 trailer images from Roboflow’s open dataset; achieved 98 % IoU ≥ 0.5 accuracy after three-tier QC.