DrillKit is an AI-powered tool that turns your online lesson chat from Zoom, Google Meet, Skype, or Microsoft Teams into professional, print-ready ESL lessons in under 30 seconds. It automatically extracts vocabulary, generates exercises (gap-fill, error correction, vocab match, word formation, and more), and produces beautiful PDFs tailored to the student's CEFR level and native language.

Does DrillKit support Zoom chats?

Yes! Simply copy the chat text from your Zoom, Google Meet, Skype, or Microsoft Teams lesson and paste it directly into the input area. You can also paste vocabulary lists or handwritten notes. DrillKit's AI strips out timestamps, sender names, and other noise automatically.

Can I export lessons to PDF?

Absolutely. Every lesson is generated as a high-quality, print-ready PDF. Pro users can also export to Google Docs and customize branding with their school logo and accent colors.

Which languages are supported?

DrillKit supports translations in 11 languages: Spanish, French, German, Italian, Portuguese, Japanese, Korean, Chinese, Arabic, Turkish, and Russian. The AI auto-detects the student's native language from the chat transcript.

What exercise types can DrillKit generate?

DrillKit generates 11 exercise types: Gap Fill, Vocab Match, Error Correction, Sentence Completion, Multiple Choice, Word Scramble, Sentence Reorder, Translation, Word Formation (Cambridge Part 3), Key Word Transformation (Cambridge Part 4), and Dialogue Builder.

How does the AI extract vocabulary from chat transcripts?

The AI scans your pasted chat transcript, strips out noise (timestamps, reactions, media placeholders), identifies teachable vocabulary, phrasal verbs, idioms, and student errors, then scores each item by relevance. You review and select the best items before generating exercises.

Is DrillKit free to use?

Yes! The free plan includes 20 generations per month with full access to all 13 exercise types, all lesson styles, and custom branding. The Pro plan ($5.99/month or $4.00/month billed yearly) unlocks unlimited generations and priority AI processing.

Can I customize the lesson with my school branding?

Pro users can add their school or institution name, choose custom accent colors, select from 5 lesson styles (Standard, Classic, Modern, Playful, Academic), and add footer text to every generated lesson.

Teaching Connected Speech in English: ESL Pronunciation Guide | DrillKit Blog

Why Students Can't Understand Native Speakers

The most common listening complaint from intermediate ESL learners: 'I understand my teacher, but I can't understand native speakers talking to each other.' The gap is almost entirely explained by connected speech.

Textbook recordings are typically spoken slowly, clearly, and with each word distinctly produced. Natural speech merges, drops, and transforms sounds at speed in ways that make it sound like a completely different phenomenon to learners trained on carefully enunciated classroom English.

The Four Key Connected Speech Features

1. Linking
When a word ending in a consonant is followed by a vowel, the consonant links to the next syllable: 'an apple' sounds like 'a-napple.' 'Come in' sounds like 'co-min.' 'Turn it off' sounds like 'tur-ni-toff.'

2. Elision (sound deletion)
Sounds that are dropped in natural speech:
• Final /t/ and /d/ before consonants: 'last week' → 'las' week,' 'good morning' → 'goo'morning'
• 'h' in weak forms: 'What does he want?' → 'What duzzy want?'

3. Assimilation (sound change)
One sound changes under the influence of a neighboring sound:
• 'ten boys' → 'tem boys' (/n/ → /m/ before /b/)
• 'eighth grade' → 'eigkth grade' (/t/ → /k/ before /g/)

4. Weak forms
Function words (to, for, of, and, but, have, can, are, was) are rarely stressed in natural speech and are reduced to weak forms using the schwa:
• 'can' → /kən/ ('I can go' → 'I c'n go')
• 'for' → /fə/ ('Wait for me' → 'Wait f' me')
• 'and' → /ən/ ('fish and chips' → 'fish 'n chips')

Connected Speech Features

🔗

Linking

Consonant + vowel linking across word boundaries — makes speech flow

✂️

Elision

Sound deletion at word boundaries — native speakers drop sounds learners expect to hear

🔄

Weak Forms

Function words reduced to schwa — the difference between 'textbook English' and real speech

Teacher Tip

“Transcribe a 30-second audio clip from natural speech (not a textbook recording) together with the student. When they can't hear a word, write exactly what they heard. Then reveal the actual words and demonstrate the connected speech feature that created the confusion. This 'decoding puzzle' approach makes connected speech features immediately concrete and memorable.”

Frequently Asked Questions

Should ESL students be taught to produce connected speech themselves?

For receptive purposes, all levels. For production, B2+ — where attempting natural rhythm and linking enhances intelligibility rather than creating distracting 'trying too hard' effects at lower levels.

Does British and American connected speech differ?

Yes. Intrusive /r/ (linking 'r' between vowels) is characteristic of rhotic varieties (American) and some non-rhotic British varieties. Glottal stops are more common in British English. Elision patterns are broadly similar; the resulting acoustic results differ somewhat.

How much time should connected speech get in a lesson?

It shouldn't be an entire lesson unless it's a dedicated pronunciation course. More effective: brief awareness moments during listening work. 'Did you notice she said 'gonna' there? That's 'going to' in connected speech. Let's listen again.'

Connected Speech in English: Teaching What Learners Actually Hear