0-to-1
Concept to live product
I handle the shape of the product, not just the implementation.
For EdTech teams, language schools, and AI-first learning products
Build AI learning products that actually survive real users.
I help EdTech teams ship learner practice, feedback, and teacher workflow tools without needing a full internal product squad. Product direction, UX, prompts, architecture, build, and launch in one loop.
What I actually bring
Voice + feedback
Practice loops people use
Speaking, correction, shadowing, and review systems built around actual learners.
Teacher reality
Workflow matters
I design for teacher time, learner motivation, and the messiness of real delivery.
Current proof
Live products, not just pilots
english-exercises.org, Firstly Academy, and WhatsEnglish give me real product feedback loops to work from.
Best for teams that need one useful thing in production fast: more learner practice, faster feedback, less teacher admin, or a stronger AI layer inside an existing product.
Why me
Not a generic dev shop.
I taught English before I started building products, so I do not have to guess where learners get stuck or where teachers lose time.
The point is not to bolt AI onto education. The point is to design practice, feedback, and workflow systems that actually improve the experience for learners and teachers. I use AI explicitly to move faster and to make the product more capable, but I keep product logic, architecture, QA, and learning design under human control.
Best current fit
Small to mid-sized teams that need a builder who can connect pedagogy, product judgment, AI workflows, and implementation without turning the project into process theatre.
- Language learning, tutoring, exam prep, and teacher tooling.
- Speaking, writing, correction, shadowing, and learner feedback systems.
- Teacher workflows, review layers, and learner follow-up automation.
Offer
Productized work, not vague capacity.
The first step should be a focused build with a clear outcome, not an open-ended pile of hours.
Fastest path
AI Pilot Sprint
Build one high-value learner or teacher workflow fast, prove it with real usage, and leave with something concrete instead of another AI deck.
- Speaking or writing practice prototype
- Voice-note correction loop
- WhatsApp or web-based micro-practice flow
Learner-facing
Practice Layer
Add meaningful practice to an existing product: speaking, writing, revision, shadowing, or feedback loops that make learners come back.
- STT to LLM to TTS voice workflows
- Rubric-based writing or speaking feedback
- Progress loops shaped by learner behaviour
Teacher-facing
Workflow Automation
Turn lesson notes, corrections, follow-up, and learner context into a usable system so teachers spend less time on admin and more on actual teaching.
- Lesson summaries and next-step generation
- Mistake capture and review dashboards
- Follow-up emails or homework workflows
Reusable system
Assessment Engine
Build scoring, feedback, and review logic into a reusable product system instead of leaving evaluation trapped in teacher intuition or scattered docs.
- Speaking or writing rubric engines
- Score explanation layers
- Teacher review and moderation workflows
Process
How I run a build.
01
Pick one narrow outcome
We start with a focused user or operational problem: more speaking practice, faster feedback, less teacher admin, or a better follow-up loop.
02
Ship something real fast
I build the first useful version end to end: product shape, UX, prompts, pipelines, architecture, QA, and launch.
03
Tighten with evidence
Once people touch it, we use the signal properly. Keep what works, fix what drags, and decide whether to expand, pause, or productize further.
Fit
Where this works best.
This is a strong fit for product-shaped EdTech problems. It is a weak fit for giant enterprise delivery theatre.
Good fit
Teams I can help quickly
- You are an EdTech team, language school, or learning business that needs one useful thing in production, not a six-month strategy exercise.
- You want someone who can work across product thinking, pedagogy, UX, AI workflows, and implementation.
- You need stronger practice, feedback, or teacher workflows but do not want to hire a full internal squad yet.
- You care about whether learners and teachers actually use the product, not just whether the demo looked impressive.
Not the best fit
Projects I would avoid
- You want a generic dev shop to build a huge custom platform from a giant requirements doc on day one.
- You want to buy vague AI innovation without a clear user or business outcome.
- You need a narrow infra or enterprise IT contractor rather than a product-shaped builder.
- You need a big delivery team with procurement paperwork before any real work starts.
Contact
If you have a learning product idea or workflow bottleneck, send it over.
The best first message is simple: what you are building, where the friction is, and what outcome you want to improve.