English
at scale

The problem

GCS sells bilingual talent. The bilingual part has a ceiling.

The promise on the site is "top bilingual professionals who deliver from day one". That adjective is what makes a professional placeable in a client-facing role — and today the company's ability to produce it depends entirely on one instructor's calendar.

The opportunity

The bottleneck isn't talent.
It's the language.

Colombia has a deep pool of capable, qualified professionals — that is the whole premise of the GCS model. What most of them don't yet have is the language level a client-facing role demands.

• 01Today "bilingual" is a hiring filter, not a capability the company builds. Strong candidates are turned away for something that can be taught, while hiring competes for the small, expensive slice of the market that already arrives bilingual.
• 02Anyone can recruit bilingual people. Very few can produce them. Competing for the same pool is competing on price. Training reliably means creating supply nobody else has.
• 03The talent is already inside the company too. People doing back-office work today who are one language level away from client-facing roles — and from the billing that comes with them.

Training language systematically turns a hiring constraint into a hiring advantage — and widens the supply of the exact thing GCS sells.

Why the current setup can't fix it

The current setup can't produce that at scale

• 01It cannot assess speaking or writing. No plugin of that generation can. Every recording and every paragraph has to be judged by a person, one at a time.
• 02Content goes in by hand, screen by screen. Material that already exists in Drive cannot be loaded in bulk.
• 03Progress isn't visible to anyone but the instructor. No cohort view, no completion reporting, no way to answer "who's on track".
• 04It's a blog with a course plugin bolted on. WordPress wasn't built for training. LearnDash adds courses; every capability past that — learning paths, gamification, reporting, memberships, certificates — is another plugin, another annual licence, another vendor that can stop maintaining it. And even buying all of them, speech and writing assessment still doesn't exist in that ecosystem.

The platform

A learning platform built for languages, owned by the company

Courses run on a main path with side lessons for the work situations the team actually deals with. Nothing unlocks because time passed — it unlocks when a learner passes.

Deployed once per audience. English for the Colombia team now, Spanish for the US team later, company onboarding after that — same software, separate installs, separate branding.

Feature · Content authoring

Content gets in without an engineer

This is the part LearnDash gets wrong and the reason the current library never made it onto the platform.

Existing material is reused, not rebuilt. Authoring stays with the instructor, not with the development team — which is what makes the platform grow without adding cost.

Feature · Automatic assessment

The platform grades speech and writing on its own

Speaking is scored sound by sound against a reference, then explained in plain language. Written answers get instant feedback. Every word carries a native recording and its phonetic transcription, pulled automatically.

This is the capability nothing on the market gives the company today, and the one that removes the human bottleneck from practice.

Feature · Teaching operations

One screen replaces chasing people over Slack

Who's moving, who's stalled, what's waiting for review. Reminders go out on their own. Courses get assigned to a whole team instead of one person at a time.

Live sessions stay exactly as they are — but the instructor walks in already knowing what to work on, instead of spending the hour diagnosing.

Feature set · Version 1

What the first version ships with

Feature set · Deliberately later

Real scope, deferred on purpose

None of this changes what gets built now. Each one waits on a decision or on the first version proving itself.

• Points and a redeemable rewards catalogueWaits on the reward rules and someone owning the catalogue
• Certificates and verifiable credentialsOnly meaningful once levels are formally assessed
• Automatic availability from Google CalendarVersion 1 books against slots the instructor sets by hand; reading real calendar availability comes later
• Multiple tutors, each with their own calendarThe programme is designed once and taught by a team — the instructor moves from delivering every class to setting the standard
• Team-wide assignment and compliance reportingBelongs to the company onboarding deployment
• Discussion and communityThe team already lives in Slack
• Spanish for the US teamSame software, second deployment

What it solves for dlivrd-GCS

Training stops being a cost and becomes capacity

What it solves for the teaching side

The instructor's reach stops being capped by the calendar

• 01Practice grades itself. Drills, recordings and repetition are handled by the platform. Review moves from every submission to the ones that count.
• 02Nobody has to be chased. Reminders and streaks do the follow-up that is currently done by hand, message by message.
• 03Existing material gets reused. Slide decks and vocabulary sets already built come across instead of being rebuilt.
• 04Live sessions get sharper. The hour is spent teaching instead of diagnosing, because the gaps are already measured.

Same method, same sessions, many more people reached.

Cost

Less than a quarter of what the company pays right now

To start

Four inputs, and two of them block everything