Product builder for language learning.

I design and ship AI-powered learner practice, feedback, and teacher workflow systems from concept to live product.

Current proof
7k weekly users on english-exercises.org
What I build
Practice, feedback, and teacher workflow systems
Why it works
Teaching insight plus explicit AI leverage
Hand-drawn portrait of Charlie Crowley

Teacher-informed product building.

Former TESOL teacher now building EdTech products end to end.

I understand the learner side, the teacher side, and the product side. That lets me shape the workflow, the UX, the prompts, the data model, and the shipped product without treating education like generic SaaS.

Current focus

Running english-exercises.org, growing Firstly Academy, and testing WhatsEnglish while building toward a connected ecosystem of learner practice and feedback products.

  • 0-to-1 product work across free tools, AI speaking practice, and WhatsApp-based learning.
  • Products shaped by classroom reality: speaking time, feedback speed, and teacher workload.
  • AI used explicitly for speed and capability, with quality control kept human.

A connected language-learning product stack.

Not one app pretending to do everything. A set of products solving different learning problems while sharing useful context.

Shared core

Shared data, shared context, shared learning signal.

The goal is simple: if a learner practises in one product, the next product should know enough to be more useful.

Firstly Academy

Live

AI speaking practice built around fluency, feedback, and repeat use.

english-exercises.org

Live

Free English practice with 7k weekly users, top-10 Google reach, and zero-login access.

Teacher's Scribe

Embedded

A teaching workflow that turns lesson recordings into usable notes, mistakes, and follow-up.

WhatsEnglish

Testing with users

WhatsApp-based AI English practice with voice correction, shadowing audio, and teacher review.

AI Voice Tutor

Embedded

Conversation practice that uses lesson transcripts and cross-product learner history.

MemEnglish

Planned

Fast mobile drills designed as a top-of-funnel practice product.

Products, not prototypes.

What I have built across learner practice, feedback, and teacher workflows.

Production

english-exercises.org

A free English practice product with 7k weekly users, search-led growth, and fast publishing systems.

Homepage screenshot from english-exercises.org
  • Next.js 15
  • TypeScript
  • SSR / SSG
  • Template system
Production

Firstly Academy

AI speaking practice for exam-focused learners, with voice feedback, progress loops, and around 40 users.

Homepage screenshot from Firstly Academy
  • React
  • Node.js
  • PostgreSQL
  • Deepgram
  • Groq
Embedded feature

Teacher's Scribe

A teacher workflow inside Firstly Academy that turns lesson recordings into reviewed mistakes, notes, and follow-up.

Teacher's Scribe workflow inside Firstly Academy
  • Node.js
  • WhisperX
  • WebSockets
  • PostgreSQL
User testing

WhatsEnglish

A WhatsApp-native AI practice product in user testing, with correction, shadowing audio, teacher review, and a production V2 backend.

Homepage screenshot from WhatsEnglish
  • Meta Cloud API
  • Node.js
  • Deepgram
  • Azure TTS
  • PostgreSQL

Product judgment first. Tools second.

01

Learning loop before feature list

I start with the learner or teacher problem, the product loop, and the data model. The interface follows the product logic, not the other way round.

02

AI as leverage, not autopilot

I use AI heavily to move faster, but product direction, standards, review, and quality control stay with me.

03

Real users over demo polish

I optimise for software that survives contact with learners, teachers, and production constraints, not just a nice prototype.

If you need a product builder for language learning, we should talk.

I’m open to product roles, pilot builds, and collaborations where teaching insight, product thinking, and technical execution all matter.