Automated, AI-powered news curation & publishing

In spring 2025, I launched Evidence Media—a fully automated pipeline that curates independent news and publishes sourced updates to X and longer, contextual articles to Substack. The goal: cut through noise, surface what matters, and make every claim traceable.

Project Overview

Evidence Media filters and aggregates high-signal stories across Business & Economics, Current Affairs, Environment, Health, International Affairs, Politics, Science & Tech, and Society. Each post cites original sources (think academic footnotes) so readers can verify quickly or dig deeper. The system is 100% automated end-to-end: discovery → analysis → drafting → sourcing → publishing.

My Role

Solo build, from vision to execution:

• Product & Editorial Strategy: mission, voice, cadence, categories, and KPIs.

• Backend & Automation: modular Python scripts orchestrated as a pipeline.

• AI Orchestration: multi-model stack (OpenAI, xAI, Perplexity) for context, summarization, and sourcing.

• Security & CI/CD: GitHub Actions for scheduling; secrets managed with HashiCorp Vault.

Stack & Tools

Python, Flask, BeautifulSoup, Selenium, Tweepy, GitHub Actions, HashiCorp Vault, plus OpenAI/xAI/Perplexity APIs, X API v2, and Google APIs. Chosen for reliability, strong docs, and automation-friendly ecosystems.

Design & UX Highlights

Content is the interface. I standardized concise, source-backed X posts and deeper, categorized Substack briefs. To reduce LLM hallucinations and bias, I used layered prompt strategies, constraints, and dynamic filtering—always prioritizing verifiability.

Deployment & Scalability

The pipeline runs from a GitHub-hosted codebase with scheduled workflows. Components (scraping, AI generation, formatting, publishing) are decoupled and horizontally scalable. Real limits are API quotas, not architecture.

Outcomes

• X account grew past 500 organic followers; Substack growth steady.

• Consistent engagement on sourced, high-signal topics.

• Stable, resilient pipeline capable of continuous publishing with minimal ops.

What I Took Away

• Deepened expertise in Python automation, scraping, prompt engineering, data integrity, CI/CD, API orchestration, and security.

• Learned to make AI helpful in editorial contexts—supporting human understanding with transparent sourcing.

• Reaffirmed a core product principle: iterate quickly, observe real-world feedback, improve continuously.

If you’re exploring AI-driven media or research automation, and want systems that are fast, scalable, and verifiable, I’d love to chat.

AI-Powered Manuscript Correction & Editorial Review

In early 2025, I built Jean The Writer, an AI-powered backend tool designed to help writers turn raw drafts into polished, publication-ready manuscripts. The project began with a unique request: correcting a 600-page French manuscript written entirely without punctuation. What started as a one-off challenge quickly evolved into a scalable system for independent writers.

Project Overview

Jean The Writer automates grammar and syntax correction, delivers chapter-by-chapter editorial reviews, and compiles clean Google Docs output with professional formatting. Unlike generic writing assistants, it’s built for long-form manuscripts, handling large texts in blocks, ensuring stylistic continuity, and even suggesting publishers. For authors, this means saving weeks of manual editing and reducing costly editorial fees, all while keeping their voice intact.

My Role

This was a solo build, and I owned the full cycle:

• Product Strategy → Designed the workflow around a client’s needs (correction, review, publisher search).

• Architecture & AI Workflows → Built modular processes with Chain of Thoughts logic and resume-capable scripts.

• AI Engineering → Selected the best LLMs (GPT-4o mini for correction/review, OpenAI Deep Search for publishers).

• Automation & Delivery → Integrated with Google Docs API for formatting (Times New Roman, 12pt, 1.5 spacing) and optional diff highlighting.

Stack & Tools:

Node.js, TypeScript, OpenAI API, Google Docs/Drive APIs, diff-match-patch, Jest, Prettier, ESLint.

Outcomes

Successfully processed 8 manuscripts (French & Spanish).

First commission: a 600-page punctuation-free novel, corrected and reviewed.

Authors praised the corrections and publisher recommendations, though editorial feedback sometimes met resistance, a valuable reminder that style is deeply personal.

What I Learned

Jean The Writer reinforced my skills as both developer and product builder. I mastered AI workflows for edge cases, advanced prompt engineering at scale, and strengthened integrations with OpenAI and Google APIs. Most importantly, I learned that AI’s role in writing isn’t to replace creativity, but to empower it, automating the repetitive so authors can focus on storytelling.

In 2024, I launched Wise Duck Dev GPTs, a platform that grew from a simple portfolio idea into one of the largest custom GPT ecosystems for developers. Today, it hosts 800+ specialized GPTs designed to support web, mobile, AI, blockchain, and video game development.

The goal was simple: help developers code faster, solve problems smarter, and stay creative by giving them access to AI-powered tools tailored to their workflows. What started as a personal library of 10 GPTs quickly scaled into a platform used by hundreds of developers, generating over 10,000 interactions in its first year.

As a solo builder, I led every stage, from product design to full-stack development, automation, and AI engineering:

• Built the platform with Next.js, TypeScript, Sass, and deployed it as a PWA on Vercel.

• Designed modular architecture with Prisma and Meilisearch for scalability and fast search.

• Automated GPT creation and bilingual SEO documentation with a 40-step Make.com workflow, saving hundreds of hours.

• Engineered 800+ GPTs with advanced prompt engineering and Chain of Thoughts logic to ensure consistency, quality, and reliability across the ecosystem.

What made this project unique is not just the scale, but the philosophy: I wanted to create a platform by developers, for developers. Every feature — from clean navigation to AI-assisted workflows — was designed to keep the experience simple, efficient, and inspiring.

Outcomes:

Version 2 launched with 400+ active users and more than 10,000 GPT uses.

Positive feedback highlighted the clarity of the interface, scalability, and the usefulness of automation-driven workflows.

What I took away: Wise Duck Dev GPTs taught me how to take a personal productivity tool and scale it into a product with global value. It strengthened my skills in automation, modular design, multilingual SEO, and AI-first development, and showed me how to align vision, technology, and user experience into one coherent platform.

👉 wiseduckdevgpts.com

In early 2024, I co-developed Family Flow, a collaborative web app built to simplify task sharing and household planning. The idea came from my own experience as a new parent: families often juggle countless responsibilities, yet lack a simple, shared space to stay aligned. Family Flow was our answer — a tool designed to reduce mental load and improve everyday communication.

As Scrum Master and Lead Front-End Developer, I helped shape the product vision, design the user experience, and build the app with React, TypeScript, Mantine UI, and Vitest, deploying it via Railway. We focused on creating an interface that felt intuitive for everyone — from parents to teenagers — through features like color-coded task types, chronological sorting, smart filters, and responsive design optimized for phones and tablets.

Delivered in just five weeks as part of O’clock’s intensive 1,500-hour Web & Mobile Development program, Family Flow was tested by real families who praised its clarity, ease of use, and calming UI. While we paused development to move on to new projects, the app was designed with scalability in mind, with future possibilities ranging from PWA deployment to AI-powered assistants.

For me, Family Flow was more than a school project — it was my first full end-to-end development experience, where I learned how to turn a personal observation into a real, functioning product. It strengthened my skills in product strategy, front-end architecture, and agile teamwork, and it confirmed something I believe deeply: the best ideas often start with real-life problems, and the best solutions are the ones that feel natural, simple, and human.