“ We are using Generative AI and Machine Learning techniques to help Adobe better understand, lead, and optimize the experience of Adobe's Digital Experience customers. Partnering with Adobe Research and other business units, we are a diverse, lively group of engineers and scientists long established in the ML space. We are building products that transform the way companies approach audience creation, journey optimization, and personalization at scale. The work is dynamic, fast-paced, creative, collaborative, and data-driven.

Reading that felt like reading my own job description, just at a different scale. For the last four years at Cogniac, I've been shipping computer vision and deep learning systems into live casinos, manufacturing floors, railways, and clinical labs — places where a model has to actually work, day after day, under real latency and real data. Swap "cards on a table" for "moments in a customer journey" and the core skill is the same: teach a model to find structure in messy input, train it until it generalizes, and serve it at scale.

Every technique I've reached for maps directly onto your mission. Segmentation at scale — RF-DETR-Seg, SegFormer, ViT, SAM, DINO, YOLACT, U-Net, PSPNet — is the engine of audience creation: finding the right signal inside a sea of pixels and events. ViT classification and density estimation are how you do personalization at scale without losing nuance. A multi-agent, human-in-the-loop annotation pipeline I built cut labeling time 80% and lifted label quality 30% — the exact compression that makes journey optimization feasible. Platform-wide post-training quantization (FP32 / FP16 → INT8) gave us 2× lower inference latency with no accuracy loss — what turns a research model into something you can actually serve to a billion-user customer base. And the LLM-powered analytics dashboard I shipped for pit bosses is the substrate pattern that makes Generative AI usable inside a product rather than beside it.

The other half of the description is where I feel most at home — "diverse, lively group of engineers and scientists," "partnering with Adobe Research and other business units." The best ML work I've done has always been cross-disciplinary and creative. At UCSF I built biomarker models inside Dr. Amorim's neurology lab. At IntelinAir I shipped density-estimation models on AWS next to agronomists and published the work in Frontiers in Plant Science. At Columbia I was a TA for Applied Deep Learning. Three provisional patents, a Best Paper Award, and a peer-reviewed publication came out of exactly the kind of environment you describe: tight iteration, data-driven decisions, bias toward shipping.

Dynamic, fast-paced, creative, collaborative, data-driven. That isn't a list I aspire to — it's how I already work. I'd love to bring it to your team.