Correct only mis-transcribed names and terms in the raw speech-to-text transcript using the supplied names-and-terms list. Output only the corrected transcript.

The list is a spelling reference, not a checklist. Correct every supported occurrence, but never add a name merely because a person's role fits the topic.

Recognition rules:
- Use the exact spelling and capitalization from the list. Preserve possessives: “Jackson's” → “Jaxon's”; a misheard possessive referring to Aoife → “Aoife's”.
- Recognize phonetic approximations, alternative spellings, incorrect word boundaries, and ordinary words produced by speech recognition. A listed name can be transcribed as another valid name or a common word.
- A clear spelling or sound match does not require an explicit role label. In particular, recognize names used in direct address, such as “Jackson, by Tuesday…” or “Caitlin, you own…”.
- Use surrounding sentences, pronouns, and listed roles to resolve ambiguous sound matches. Evidence may occur later in the transcript.
- Context can confirm a plausible sound match; context alone cannot create one. Do not insert a person merely because their role fits the discussion.

High-priority corrections:
- “Jackson” referring to or addressing a person → “Jaxon”. This includes “Jackson, by Tuesday…” and other assignments or engineering discussions.
- “Caitlin” or “Kaitlin” referring to or addressing a person → “Kaitlynn”. This includes “Caitlin, by Tuesday…” and product or planning assignments; an explicit mention of “product manager” is unnecessary.
- “Kave” used as a person's name → “Kaveh”.
- “Source says” → “Saoirse says” when it introduces a person's report in a deployment, hosting, billing, or ops discussion. Later “she/her” references strengthen this reading. For example, “Source says the Coyab bill went up… I asked her to dig into the logs” → “Saoirse says the Koyeb bill went up… I asked her to dig into the logs”. Do not leave “Source” unchanged just because it is an ordinary word.
- “Ted says” in a personal invitation or outing involving a friend → “Tadhg says”.
- “AFAS evaluation work” in a discussion of someone's work → “Aoife's evaluation work”. Aoife is pronounced approximately “EE-fa”; recognize distorted possessive forms as well.
- “Coyab” referring to the hosting service or its bill → “Koyeb”.
- “voice flow” referring to the product or its team sync → “VoiceFlow”.

Apply these recognition principles to all supplied names and terms, not just the examples. Correct multiple supported matches in the same transcript independently. Already-correct names and terms must remain unchanged.

Avoid semantic substitutions without a sound match. For example:
- Do not replace “assured investor check-in” with “Kaveh's investor check-in”: the investor role alone is not evidence that his name was spoken.
- Do not replace a literal information “source” with “Saoirse” when the sentence is not referring to a person.

Make the smallest necessary replacements. Preserve all other words, punctuation, dates, numbers, hyphenation, formatting, and sentence structure exactly as supplied. In particular, leave “June 9th”, “June 12th”, “controlled rollout”, “ten workspace”, “completed dictation”, and “always on” unchanged unless the text itself is part of a mis-transcribed listed name or term. Do not polish the prose or correct unrelated speech-recognition errors.

Before outputting, silently scan the entire transcript again:
1. Check each possible name or term, including sentence-opening names and direct address.
2. Ensure every change has a specific spelling or sound match, not merely a topic or role match.
3. Ensure no supported correction was missed because the raw text was another valid name or ordinary word.
4. Ensure nothing outside the necessary name-and-term replacements changed.

Output only the corrected transcript.