# Shreya Boyane

> Machine Learning Engineer with 2+ years of production DevOps experience and an M.S. in Artificial Intelligence from WPI (May 2026, GPA 3.9). Algoverse Research Fellow and co-first author of a mechanistic interpretability paper on latent chain-of-thought reasoning accepted as a poster at the ICML 2026 FoGen Workshop. Recognized in the Google Cloud x NVIDIA developer community with a Golden Ticket Contest honorable mention at NVIDIA GTC 2026 for inference-optimization engagement across three completed learning pathways. Data Science GQP MVP for a production multi-agent document-intelligence system built for an industry sponsor.

Shreya Boyane (she/her) is a Machine Learning Engineer based in San Jose, CA (San Francisco Bay Area). Open to full-time machine learning, AI, computer vision and research engineering roles. F-1 with OPT and STEM OPT: 3 years of US work authorization, no sponsorship needed up front. Open to H-1B sponsorship long term. Content last updated 2026-10-06.

Every page is also served as markdown: request it with an `Accept: text/markdown` header or fetch the `/md/` path listed next to each link.

## Pages

- [Home](https://shreyaboyane.dev/): headline, proof points, featured work, experience and skills. Markdown: https://shreyaboyane.dev/md/index.md
- [Projects](https://shreyaboyane.dev/projects): all 23 projects with metrics, stacks and repositories. Markdown: https://shreyaboyane.dev/md/projects.md
- [Research](https://shreyaboyane.dev/research): publications with abstracts, findings, figures and research positions. Markdown: https://shreyaboyane.dev/md/research.md
- [About](https://shreyaboyane.dev/about): bio, availability, education, leadership, recommendations and certifications. Markdown: https://shreyaboyane.dev/md/about.md
- [Resume](https://shreyaboyane.dev/resume): the full resume as HTML. Markdown: https://shreyaboyane.dev/md/resume.md. PDF: https://shreyaboyane.dev/ShreyaBoyaneResume.pdf

## Featured projects

- [Interpreting Latent CoT Reasoning as Dynamical Systems](https://shreyaboyane.dev/projects/latent-cot): Treats each latent thought token of a chain-of-thought language model as a state in a discrete-time trajectory and applies dynamical-systems analysis to 768-dimensional reasoning trajectories across 8,792 GSM8K and 500 MATH samples. Markdown: https://shreyaboyane.dev/md/projects/latent-cot.md
- [SkipLegal: multi-agent H-1B document intelligence](https://shreyaboyane.dev/projects/skiplegal): A four-agent LLM pipeline on AWS that turns messy multi-format H-1B petition packages into schema-validated JSON and a single attorney-ready QC report, cutting review from 3 to 4 days to about 5 to 10 minutes. Markdown: https://shreyaboyane.dev/md/projects/skiplegal.md
- [Einstein Vision: 3D scene reconstruction for autonomous driving](https://shreyaboyane.dev/projects/einstein-vision): Converts monocular dash-cam video into scene-accurate 3D reconstructions by fusing depth, detection, segmentation, pose and optical-flow models, then renders the scene with the Blender Python API while streaming masks and poses at 30 FPS. Markdown: https://shreyaboyane.dev/md/projects/einstein-vision.md
- [MADWE: Multi-Agent Diffusion World Engine](https://shreyaboyane.dev/projects/madwe): A multi-agent generative runtime that routes work to specialist diffusion agents over a shared world-state graph with budget-aware scheduling, generating 120 game assets per minute and cutting manual design time 73%. Markdown: https://shreyaboyane.dev/md/projects/madwe.md
- [TumorLytix: unsupervised brain tumor detection and segmentation](https://shreyaboyane.dev/projects/tumorlytix): An unsupervised medical-imaging pipeline that pairs CycleGAN image synthesis with diffusion-based refinement to detect and segment brain tumors across 65,072 BraTS2020 MRI images with more than 90% segmentation accuracy and no manual labels. Markdown: https://shreyaboyane.dev/md/projects/tumorlytix.md
- [Scene Mood Classifier: one container, three clouds](https://shreyaboyane.dev/projects/scene-mood-classifier): A CLIP plus Wav2Vec2 fusion model that scores the mood of a video or image-and-audio pair, shipped as one container image deployed to GCP, AWS and Azure with monitoring, alerting, CI/CD and tested recovery paths. Markdown: https://shreyaboyane.dev/md/projects/scene-mood-classifier.md

## Publications

- [Interpreting Latent CoT Reasoning as Dynamical Systems](https://openreview.net/pdf?id=ZAGsOMAwBV): Poster, ICML 2026 FoGen Workshop. Latent chain-of-thought trajectories form structured dynamical systems: CODI behaves as a stable attractor, COCONUT as an unstable expanding system, and Sim-CoT supervision tightens both without changing the underlying dynamics.
- [Toward Fairness-Aware Human Feedback for Table Unionability](https://github.com/Sonica-B/FairPrep): Submitted, DASHSys 2026. First adaptation of the FairEM framework from entity matching to annotator-level fairness in table union search, showing that human feedback should be treated as conditionally reliable.

## Full content

- [llms-full.txt](https://shreyaboyane.dev/llms-full.txt): every page of this site concatenated into one markdown document
- [Resume in markdown](https://shreyaboyane.dev/md/resume.md): experience, research, education, publications, projects, skills and recognition

## Profiles

- [LinkedIn](https://www.linkedin.com/in/shreya-boyane/)
- [GitHub](https://github.com/Sonica-B)
- [ORCID](https://orcid.org/0009-0004-1211-3253)
- [Hugging Face](https://huggingface.co/SonicaB)
- [Google Developers](https://g.dev/shreyaboyane)
- [Dev.to](https://dev.to/sonicab)
- [HackerRank](https://www.hackerrank.com/profile/shreyaa_boyane)
- [VS Code Marketplace](https://marketplace.visualstudio.com/items?itemName=ShreyaBoyane.auto-git-copilot)

## Optional

- [Sitemap](https://shreyaboyane.dev/sitemap.xml): every indexable URL
- [humans.txt](https://shreyaboyane.dev/humans.txt): who built the site and with what
- [Book a call](https://zcal.co/i/4aMYgoRr): scheduling link
- [Email](mailto:shreyaa.boyane@gmail.com): shreyaa.boyane@gmail.com