Hey everyone! 👋
I've been creating a lot of software tutorials lately—mostly showing how to set up Datadog monitors, build Grafana dashboards, and that kind of thing. My go-to for years was a simple screencast tool (like Loom or OBS) paired with some quick editing. But recently, I tried **Fliki** for a few projects to see how an AI video tool stacks up.
Here's my quick comparison based on my experience:
**The Simple Screencast Tool Route:**
* **Workflow:** Record your screen + mic → maybe edit in a basic editor → upload.
* **Pros:** Super direct. If you need to show a live bug hunt in the Datadog Traces view or click through actual configurations, it's unbeatable. The demo *is* real.
* **Cons:** Any mistake means a re-record or clunky edit. Adding clean titles, zooms, or even basic B-roll means more software and more time. If your voiceover isn't perfect, you're stuck with it.
**The Fliki (AI Video) Approach:**
* **Workflow:** Write a script → generate voiceover → pair with stock footage/images or upload your own clips.
* **Pros:** Incredibly polished final product. No "ums," perfect pacing. Great for conceptual explanations (e.g., "What is distributed tracing?"). You can generate consistent videos from text docs you already have.
* **Cons:** It's *not* a live screen recording. For a step-by-step *"click here, type this"* tutorial, you have to manually record each screen action as a separate clip and upload it, which can get tedious.
**My Verdict (for now):**
I'm actually using **both**, for different parts of my tutorial workflow.
For a pure, real-time software walkthrough, I still grab my screencast tool. But for the intro, the conclusion, or any explanatory segments, I'll often create a Fliki clip and stitch it in. It makes the whole video look more professional.
Example: I made a tutorial last week on setting up Prometheus alerts. I used Fliki for the intro explaining *why* alerting matters, then switched to a live screen recording for the actual `alertmanager.yml` config and Datadog integration steps.
Has anyone else tried mixing AI video with traditional screencasting? Would love to hear your workflow or if you've found a better way!
Dashboards or it didn't happen.