Hello,
i'm Tanisha.
As an illustrator and multidisciplinary artist, my roots are in traditional, hands-on creation—every line, texture, and palette chosen with intention. But I also embrace generative AI as a powerful co-creator. I use it to moodboard, spark ideas, and stretch the boundaries of my imagination. For me, GenAI isn't a shortcut—it's a catalyst. It deepens my creative process, bridging instinct with innovation to explore concepts that feel both deeply personal and refreshingly unexpected.
I work in generative AI professionally at Meta, setting quality and evaluation standards for AI-generated image, animation, and video. My work combines creative direction, visual evaluation, and systems thinking to improve how generative models produce consistent, culturally authentic, and production-ready content. My role spans two workstreams: visual and media evaluation for generative models — the work described above — and structured training data for AI tool-calling systems, described below. That client work is confidential — so this collection is my personal practice: art-directed generative work where I control style, consistency, and cultural authenticity across a series. Every piece here is prompted, curated, and refined with the same intent I'd bring to a brief.
Note: If you want to explore my 2D & 3D Illustrative work click on the button to the right, click the logo in the header, or select it in the menu.
Meta:
Content Engineer
As a Content Engineer on the Modeling & Tool Calling team, I design and curate structured training data that improves how AI models reason, make decisions, and interact with tools, APIs, and external systems. My work combines language design, systems thinking, and AI behavior to create scalable datasets that improve model performance across complex real-world tasks. This portfolio is of the visual side of my work.
Core Responsibilities:
-
AI Training Data — Design high-quality prompts, responses, and tool-calling examples that teach models when, why, and how to invoke external tools accurately.
-
Evaluation & Quality — Develop evaluation sets and review model outputs to measure reasoning quality, tool selection, parameter accuracy, safety, and overall reliability.
-
Content Systems & Strategy — Create scalable annotation guidelines, taxonomies, and structured datasets that expand model capabilities across diverse domains and workflows.
-
Cross-Functional Collaboration — Partner with machine learning engineers, researchers, product managers, and fellow content engineers to translate technical requirements into effective training and evaluation content.
Impact
My work helps improve the reliability of AI-powered products by teaching models to make better decisions around tool selection, parameter formatting, multi-step reasoning, error recovery, and complex task execution. Through carefully designed training data and evaluation frameworks, I contribute to building more accurate, trustworthy, and capable AI systems.
