Boniface Saruni

I specialize in the quality layer of the AI lifecycle. By combining Human-in-the-Loop (HITL) QA with multimodal annotation, I help AI teams achieve 99%+ ground-truth accuracy for Autonomous Systems and Large Language Models (LLMs). I work with 3D LiDAR, video, and images to ensure precise labeling for autonomous driving and perception projects.

I audit and refine text datasets to improve model alignment, support RLHF, and reduce hallucinations. I design audit frameworks that turn noisy data into high-signal training sets. I build reliable systems for data quality, mentor annotation teams, review datasets for accuracy and ethical alignment, and ensure AI outputs are scalable, responsible, and deployment ready.

Passionate about partnering with forward-thinking AI companies to solve complex data quality challenges.

LiDAR Annotation Specialist at Digital Divide Data (DDD) (2026-01 – Present)

• Annotate and segment 3D LiDAR point cloud data for autonomous driving projects, ensuring precise object labeling and spatial accuracy.
• Perform structured quality checks, maintaining >95% annotation accuracy and consistency across complex 3D datasets.
• Identify and document edge cases, occlusions, and ambiguous objects to support guideline refinement.
• Escalate technical inconsistencies and sensor anomalies to supervisors.
• Participate in calibration sessions and integrate feedback to improve inter-annotator consistency.

Video & Text Data Annotation Specialist at Atlas Capture (2025-11 – Present)

• Annotate complex video sequences involving object tracking, action recognition, and temporal event labeling.
• Review and label textual data for LLM evaluation tasks, ensuring context relevance and model alignment.
• Handle challenging scenarios such as occlusion, motion blur, and ambiguous samples while maintaining high consistency.
• Deliver production-level annotations aligned with AI and LLM training requirements.

Annotation and Platform Specialist at Humans in the Loop (2024-05 – 2025-08)

• Designed and delivered AI annotation training programs for Image, Video, Text (LLM), and 3D LiDAR datasets.
• Led Human-in-the-Loop workflows and reviewed pilot projects to ensure dataset accuracy and model-ready quality.
• Mentored annotators and optimized training platforms and annotation guidelines.
• Contributed to roadmap planning and process improvements for scalable AI data operations.

Project Data Annotator at Label Your Data (2024-04 – 2024-12)

• Reviewed and validated multimodal datasets (Image, Video, Text, 3D LiDAR) to ensure high accuracy and model readiness.
• Collaborated with engineers and data scientists to identify dataset gaps and refine annotation guidelines.
• Maintained detailed documentation and quality metrics throughout projects.

Delivery Project Assistant at CloudFactory (2023-04 – 2024-04)

• Coordinated AI/data annotation projects, ensuring timely delivery of high-quality datasets.
• Monitored project progress, resolved bottlenecks, and supported team training and onboarding.
• Drafted reports and aligned deliverables with client AI/ML objectives.

Data Annotation Specialist at CloudFactory (2021-01 – 2023-03)

• Reviewed and annotated diverse datasets including Images, Videos, and 3D LiDAR with high accuracy.
• Conducted QA checks, peer reviews, and contributed to guideline development.

Data Annotator at Enlabeler (2020-05 – 2021-06)

• Delivered accurate annotations across Image, Video, Text, and 3D LiDAR datasets.
• Collaborated with data scientists to meet project standards and model requirements.

AI Trainer at Sama (2018-11 – 2020-02)

• Delivered high-quality annotations and mentored junior annotators.
• Developed annotation guidelines and best practices.

Bachelor of Science in Information Technology (Computer Science) – Jomo Kenyatta University of Agriculture and Technology (JKUAT) (2015-09 – 2019-11)