Allan Kimosop Kipkemboi

Data Scientist & AI Specialist | RLHF | LLM Evaluation | NLP & Computer Vision | Enterprise Data Pipelines

View My Work Download Resume

About Me

I’m a data scientist and AI specialist with 5+ years of hands-on experience across the full AI data lifecycle. I specialize in boosting model performance through high-quality data pipeline design, human-in-the-loop systems, and RLHF/RLAIF optimization. My work spans NLP, computer vision, and multilingual data processing for enterprise-scale AI. Currently pursuing a B.Tech in Chemical Engineering while delivering production-ready AI solutions remotely.

Languages

English C2
Swahili C2
German B2

Core Competencies

Methods & Specializations

Machine LearningData ScienceNLPComputer VisionLLM EvaluationRLHF / RLAIFPrompt EngineeringModel Training/EvaluationFeature EngineeringData PipelinesStatistical Analysis

Technical Stack

PythonPandas / NumPyScikit-learnSQLC++TensorFlowPyTorchHugging Face TransformersJSON / XMLAdvanced Excel

Tools & Platforms

Scale AILabelboxCVATAppenRemotasksSurge AIAmazon MTurkDocker

Professional Experience

Generalist Data Annotation Expert (LLM Evaluation & AI Training)

Mercor (Remote) | Nov 2025 – Present
  • Led production deployment of a new RLHF data pipeline, integrating 10,000+ human-in-the-loop signals weekly, reducing user-reported factual errors by 15%.
  • Architected standardized LLM evaluation & ranking processes, improving human-alignment scores, quality output, and dataset annotation accuracy by 20%.
  • Strengthened pipeline efficiency using standardized taxonomy & QA workflows across large-scale AI training datasets (JSON/XML).

Data Developer (Machine Learning & NLP Specialist)

RWS Group (Remote) | Sep 2025 – Present
  • Architected feature engineering layer in Python/Pandas, boosting multilingual NLP dialect classification accuracy by 9% across 10+ dialects.
  • Increased CV model precision/recall via pixel-level semantic segmentation & bounding box annotation (CVAT, Labelbox).
  • Optimized evaluation workflows using pairwise comparison testing, accuracy, and F1-score validation.

Data Analyst

CloudFactory (Nairobi) | Sep 2022 – Dec 2023
  • Delivered data-driven insights using Python & SQL, improving reporting accuracy and turnaround time.
  • Designed validation workflows, data cleaning, and preprocessing pipelines to ensure data integrity.
  • Streamlined KPI tracking & client query management through optimized reporting processes.

AI Data Annotator

Remotasks (Remote) | Jun 2021 – Aug 2022
  • Scaled supervised learning datasets with image, video, and LiDAR annotations.
  • Improved speech recognition accuracy via multilingual audio transcription & segmentation.
  • Enhanced object detection performance using bounding boxes, landmarks, and 3D point clouds.

Featured Projects

ResNet-50 Image Classifier

Built and trained a PyTorch-based ResNet-50 model on a 100k-image proprietary dataset for object detection.

✅ 94% Top-1 Accuracy
🔍 View Project 🐙 GitHub

BERT Sentiment Analysis Pipeline

Fine-tuned a Hugging Face BERT model on a 500k-document corpus for real-time data insights.

✅ 92% F1-Score
🔍 View Project 🐙 GitHub

Full-Stack Data Analysis Pipeline

Engineered an end-to-end Python/Pandas/SQL pipeline generating structured insights and automated KPI reports.

✅ Streamlined Reporting & Tracking
🔍 View Project 🐙 GitHub

Let's Build AI That Works

Open to AI/ML, Data Engineering, and LLM Evaluation roles worldwide.

📧 Email

Alkim8300@gmail.com

📞 Phone

+254 745 956 337

📍 Location

Nairobi, Kenya (Open to Remote/Relocation)