Elena R. is an equity analyst at a long-only consumer fund based in Milan. She covers eighteen European consumer staples companies, and before earningscalls.dev she spent a significant portion of every earnings season doing what she calls "data janitorial work": reformatting transcripts, pulling out numbers and building comparison tables by hand. We talked about what changed once the transcripts came in as structured JSON, and why she now reads the Q&A sections she used to skim.
Interview recorded on June 5, 2026. Plans and prices mentioned are the ones the interviewee was on at the time; current plans are on the pricing page.
Elena, your quote mentions that every company comes back in the same structure, saving hours of cleanup. Can you give us a concrete before-and-after?
Before, I had a template, an Excel model where I tracked revenue guidance, margin commentary, and capex plans across my coverage universe. Populating that template meant pulling each company's transcript, finding the relevant passages, and manually copying the numbers and quotes into the right cells. With eighteen companies, that was six to eight hours per earnings season, minimum. And that was just the data entry. If a transcript had formatting issues, like page headers mixed into the text or speaker names buried in paragraphs, I'd spend extra time cleaning before I could even start.
Now the API returns flat JSON with speaker segments pre-segmented and tagged. I run a script that extracts the executive segments, searches for the specific metrics I track, and populates my template automatically. The whole process takes about fifteen minutes.
What do you do with the executive versus analyst segmentation specifically?
Two different analyses. For management commentary, I isolate executive segments and run a sentiment comparison against previous quarters. Are they using more hedging language? Are they more definitive about guidance? For the Q&A, I isolate analyst segments and look at topic clustering. When five analysts ask about the same thing, say promotional intensity or input costs, that tells me what the Street is worried about. That's signal I can use to frame my own questions before the next call.
You mentioned the free reading tier. How does that fit into your workflow?
It's my sanity check layer. When a company outside my coverage reports and I just need a quick read, I don't fire up the API. I go to the site, open the transcript, and use browser search. No login, no friction. I can send a link to a portfolio manager and they can read it immediately. That sounds trivial, but the absence of a login wall matters when you're sharing research across a team.
What was the hardest part of trusting the structured output?
The first month, I double-checked everything. I'd pull a segment from the API and compare it to the original transcript on the company's IR page. I did that maybe ten times, and every segment boundary was correct, every speaker tag matched. After that I stopped checking. The data quality is consistent enough that you can build on it without paranoia.
You're on Pro at $24.99. Is that the right tier for an individual analyst?
For my usage, absolutely. I'm not doing bulk backfills or building a database. I query specific calls during earnings season. The 3,000 requests cover that comfortably. If I were running a quant pipeline I'd need more, but for fundamental analysis, Pro is more than enough. The cost is less than my monthly coffee budget, and the ROI is measured in hours saved per quarter.
What's your biggest frustration with the platform?
Coverage for European small-caps. My coverage includes a couple of Italian and Spanish mid-caps, and one of them doesn't always have transcripts available. That's how transcript availability works for smaller issuers, but it's worth knowing before you sign up. For large and mid-caps across major European markets, I haven't had issues.
How has this changed your output quality?
I read the Q&A sections more carefully now. Before, I'd skim them because I was exhausted from the data assembly. Now I have the bandwidth to actually engage with what management is saying under pressure. The good stuff is in the unscripted moments, and before I was too tired to notice them.
Would you recommend it to other fundamental analysts?
Without hesitation. If you're covering more than ten companies and you're still doing transcript work manually, you're missing things. And with every company in the same JSON structure you can finally compare them side by side, which never really worked when I copied things out by hand.
Want to try the same workflows? You can search a whole sector, pull speaker-tagged transcripts as JSON or connect Claude through the MCP server. Plans and coverage are on the pricing page, and the full reference is in the API docs.