SEO & AI Initiative
Brainly — Product Manager
Faced with Google’s SGE update threatening organic traffic, Brainly needed a scalable strategy to defend SEO rankings and maintain traffic acquisition.
Brainly’s organic search traffic was at serious risk due to Google’s new AI search features surfacing direct answers instead of traditional search listings.
The goal was to generate SEO-optimized topical content at scale to defend and grow Brainly’s visibility while minimizing development and operational costs.
- Led the exploration of scalable SEO content generation strategies using AI and scraping tools.
- Designed a Python-based system to orchestrate topical content creation across up to 1 million subjects.
- Built AI prompting frameworks for content generation aligned with educational standards and SEO best practices.
- Negotiated and aligned cross-functional teams (AI, Content, Product, Engineering) around a lean MVP plan, cutting proposed costs drastically.
- Supervised the web scraping and QA process to ensure topical accuracy and coverage.
- Reduced projected project costs from $300K to $15K by designing in-house MVP approach.
- Achieved 93% topical coverage and 70% QA accuracy across generated content.
- Delivered full MVP framework and initial data generation plan ahead of stakeholder deadlines.
Django-Based Collaborative Dissertation Platform
Freelancer — Systems Architect
Traditional academic supervision tools (Word, email, PDFs) were unstructured, hard to track, and inefficient for managing feedback on a 300+ page dissertation under constant revision.
The goal was to build a purpose-built collaboration platform for real-time document review between PhD student and advisor.
Academic advisors needed a structured way to comment on specific sections, track changes over time, and reference data tables and figures — none of which was well-supported by traditional tools.
The opportunity was to build a lightweight but powerful Django-based app that enabled inline feedback, review tracking, and custom integration of code-generated content.
- Designed a custom collaboration platform tailored to the academic supervision workflow, enabling structured feedback and version control across long-form documents.
- Built a local-first system that prioritized data privacy, modularity, and full author control over content and revisions.
- Developed flexible editing and commenting capabilities to support granular, sentence-level review and iterative improvement.
- Implemented dynamic content handling to support integration of external elements such as data tables and figures generated in Python.
- Created administrative tools to manage documents, track feedback cycles, and support academic process transparency.
- Future-proofed the architecture to accommodate citation logic, content compilation, and extended use beyond a single dissertation.
- Eliminated friction in PhD feedback loops — supervisor could leave comments per sentence, reducing back-and-forth confusion.
- Enabled centralized review across dozens of chapters, interviews, figures, and datasets.
- Significantly reduced cognitive load for managing revision cycles and integrating comments.
- Demonstrated feasibility of a structured alternative to traditional word processors for academic research.