AI Product Engineer

Alex Shuvalov

I build document-intensive AI workflows with structured output, source traceability, and human review.

Hands-on engineering, product research, and grant-backed development.

GitHubLinkedInEmail

Selected Work

01 / Products & experiments

Featured / Industrial AI

Procedra

Explore the interface →

From an operational request and reference documents to a structured instruction draft that an expert can review.

Procedra application showing task input alongside a generated instruction draft
Actual application, synthetic demo content. Screens show the Russian-language interface and deterministic generation mode.

I designed and developed the workflow end to end: document retrieval, structured output, source context, editing, review, version history, and execution checklists.

A controlled demonstration prototype, not a production deployment. Generated instructions require qualified expert review. Application source remains private.

Python / FastAPI / Pydantic / SQL / Retrieval / Evaluation

Product case studyPresentationWorking paper

Healthtech / Product development

MedAI

Co-founder experience across product architecture, analytics, partner communication, and grant work. Listed as an author of the registered software.

Read the case study →

Document AI / Computer vision

Engineering Drawing OCR

Contributed preprocessing and recognition experiments to a larger technical-drawing project. Public reference code includes a runnable synthetic preprocessing demo and tests.

Explore the code →

Open Source

02 / Accepted contributions

Mellea

Hardened release-dispatch input validation and literal payload handling.

Merged pull request →

Backlot

Improved HubSpot object-type compatibility, GitHub pagination, and API version-header handling.

Merged pull request →

Agent Swarm

Corrected environment-flag parsing for integrations and provider configuration.

Merged pull request →

Autosubmit

Documented YAML configuration loading and override order.

Merged pull request →

Browse accepted contributions →

Background

03 / Product thinking, engineering execution

Experience across healthtech, industrial AI, and early-stage digital products: market research, competitor analysis, MVP definition, partnerships, and grant applications.

Currently pursuing a master's in Computer Science and Engineering through the Digital Ural program, following a bachelor's in Architecture and Civil Engineering.

Engineering
Python, FastAPI, Pydantic, SQL, JavaScript / TypeScript, Docker, REST APIs
Applied AI
Retrieval, LLM integration, structured output, OCR, evaluation
Quality
pytest, mypy, Ruff, GitHub Actions, validation, audit trails
Product & research
Discovery, competitor research, MVP scoping, grant-backed execution