Kavata Mutia

Analytically minded Data Specialist with 3+ years of hands-on experience in AI data annotation, content quality review, text evaluation, and structured data labeling. I have contributed to AI training pipelines by rating model outputs, writing and validating prompts, classifying text and images, and ensuring datasets meet the quality benchmarks platforms like Outlier, Remotasks, and RWS Group rely on. My background in Applied Statistics means I bring genuine rigor to tasks that need it — I don't just label fast, I label right.

I work fully remotely, manage my own time well, and genuinely enjoy the kind of careful, detail-first work that makes AI systems smarter.

Data & Content Quality Specialist - Synapse Analytics Solutions - Nairobi, Kenya (Hybrid: 3 days remote / 2 days on-site)

(2025-02)

• Reviewed and quality-assured AI-generated text outputs across multiple content verticals — evaluating responses for factual accuracy, tone consistency, and adherence to client style guides, flagging issues using structured feedback forms.
• Developed and maintained a content classification taxonomy covering 40+ topic categories, used to label and route incoming data tasks to appropriate review queues — reducing misclassification rates by 18%.
• Performed systematic data validation on structured and semi-structured datasets, applying rule-based checks and manual review to catch formatting errors, duplicates, and missing fields before client delivery.
• Wrote detailed evaluation rubrics and annotator guidelines for new project onboarding, enabling a team of 8 reviewers to reach consistent quality benchmarks within the first week of each engagement.
• Attended weekly remote stand-ups via Zoom and maintained asynchronous task updates in Trello, demonstrating reliable communication and time management across a distributed team environment.
• Produced weekly quality reports in Google Sheets summarizing error rates, reviewer performance, and dataset completion status — shared with project managers and used to inform task prioritization.
• Managed multiple concurrent review tasks independently, consistently meeting daily output targets while sustaining accuracy rates above 96% as measured by internal QA audits.

Junior Data Analyst - Institute of Public Finance (IPF) - Nairobi, Kenya (Hybrid)

(2023-05 - 2025-01)

• Analyzed and validated large public-finance datasets in Google Sheets using pivot tables, VLOOKUP/XLOOKUP, conditional formatting, and data-validation rules — ensuring integrity of figures distributed to government stakeholders.
• Produced structured Google Docs reports and briefing notes for both technical and non-technical audiences, applying plain-language principles and consistent formatting standards.
• Built Google Slides decks that translated complex statistical findings into clear visual narratives; presented findings directly to senior managers and external partners on four occasions.
• Developed and maintained detailed SOPs for the team's data collection, review, and reporting cycles, cutting average turnaround time by approximately 20%.
• Supported annotation and structured labeling of policy datasets — experience that maps directly to AI training data workflows.
• Troubleshot formula errors, import inconsistencies, and formatting bugs across the Google Workspace environment, acting as the informal Sheets resource for junior colleagues.

Statistical Research Assistant (Part-time) - Rongo University — Department of Statistics - Rongo, Kenya (On-site)

(2022-09 - 2023-04)

• Assisted faculty in collecting, cleaning, and coding primary survey data for two peer-reviewed research projects on agricultural household income and public-health outcomes.
• Transcribed and coded qualitative interview data into structured spreadsheets, applying consistent coding schemas — directly analogous to annotation work on AI training datasets.
• Ran descriptive and inferential statistical analyses in R and SPSS; produced visualizations and summary tables included in final research reports.
• Cross-checked data entries for accuracy and flagged anomalies for senior researchers, maintaining data integrity across datasets of 2,000+ records.
• Prepared and formatted references, citations, and appendices for three academic manuscripts submitted to regional journals.

Bachelor of Science - Applied Statistics with Computing - Rongo University (2020 - 2024)