Valentina Akpan — Founder, Rellatech
The brief was simple. Fourteen months of weekly coaching clinics were sitting inside Otter.ai, transcribed but unusable. They needed to be in one place, ordered newest to oldest, ready to be searched, referenced, and reused inside the business.
Below is what the finished system looks like, the numbers behind it, and why this kind of cleanup matters far more than it sounds.
The Outcome
A single Word document, ordered newest to oldest, containing every coaching session in clean, readable form. From login to delivered file, the working pipeline ran in around ten minutes, unattended. No copy and paste, no manual exports, no cleanup afterwards.
51
Transcripts compiled
2.3M
Characters captured
10 min
Unattended runtime
0
Parse errors
Why It Mattered
Coaching businesses live and die by what they remember. Every clinic call had decisions, frameworks, and client breakthroughs locked inside it. Sitting one by one inside Otter.ai, that knowledge was effectively invisible. Nobody was going to scroll through 51 separate recordings to find the one quote, the one explanation, the one piece of advice that would unlock a client.
In a single document, that same content becomes a true reference library. Searchable in seconds. Quotable inside emails, proposals, and content. Easy to hand off to a team member, an editor, or a future ghostwriter.
Knowledge that lives in 51 separate recordings is not a knowledge base. It is a graveyard.
What Actually Got Built
The final piece is a small Streamlit app that lives on the desktop. The owner enters their Otter.ai login, the app fetches every transcript in order, and a polished Word document lands on the desktop. That is the entire user experience.
Underneath, Python is doing the heavy lifting through the open-source otterai-py library and python-docx. The result is a tool the business now owns and can run again at any time, with no logins to remember, no manual steps, and no monthly subscription attached.
The Bigger Lesson
This was never really about Otter.ai. It was about a pattern I see in almost every business I work with. There is enormous value sitting inside the tools you already pay for. Recordings, customer notes, support emails, sales calls. Most of it is locked behind interfaces that were never designed for retrieval.
A clean extraction, the right format, a small piece of automation, and suddenly the same content becomes a strategic asset. That is the work I love doing.
Have a Pile of Data You Cannot Use?
If your business is sitting on transcripts, recordings, exports, or notes that nobody has the time to organise, this is exactly the kind of project I take on. Quiet, unglamorous, and genuinely useful.
Book a Consultation