From Brief to Video: Inside Dobidy's AI Workflow

Most video ad tools start with a blank form: product name here, description here, pick a template. Dobidy starts with a conversation. Here's how a short chat and a few product photos become finished 10-second video ads — step by step, using Dobidy's Omni Tool.

Step 1: Discovery — A Conversation, Not a Form

When you start a new project, you chat with Abil, Dobidy's discovery agent. Instead of staring at empty input fields, you answer targeted questions in plain language:

• "What are you advertising?"
• "Who's your ideal customer?"
• "What platforms do you want to run this on?"
• "What's the one thing viewers should take away?"

Abil extracts the nine data points that drive effective video ads: product name, product description, target audience, key selling points, tone, platform, call-to-action, brand colors, and specific requirements. The conversation takes about two minutes. If you've already set up a brand profile with your product details, Abil skips the basics and jumps straight to what's unique about this particular ad.

Once Abil has enough, it generates a Creative Brief — a structured document summarizing everything the downstream AI agents need. You review and approve it right in the chat. If something is off, tell Abil what to change and it regenerates. No separate approval page, no context switching.

You also upload product photos at this stage: up to three product angles plus a logo. These images aren't decorative — they become reference elements that the video model uses to render your actual product in the final video.

Step 2: Three Scenarios, One Pick

With the brief locked in, Bisli — the scenario agent — generates three distinct ad concepts for your product. Each scenario is a complete creative direction:

• Concept overview — the strategic angle (humor, aspiration, problem-solution, social proof)
• Hook — the first-second visual or line that stops the scroll
• Scene description — what the viewer sees
• On-screen text — the words overlaid on the video
• Spoken line — a voiceover line (25-30 words, timed for 10 seconds)
• Call-to-action — what you want the viewer to do next

These aren't vague taglines. They're fully realized ad concepts. One might lead with humor, another with a bold product claim, a third with aspirational lifestyle imagery. Same product, three angles — you pick the one that fits your brand.

When you hit Continue, Carmen — the video director agent — transforms all three scenarios into complete production prompts, one per scenario. Carmen specifies timestamped scenes with exact camera movements, lighting, and element placement. Your product photos are mapped to specific moments using @Element tags, so the video model knows exactly when and where to feature your product.

These prompts are purpose-built for Kling O3, the video generation model that produces the final output. Every camera angle, transition, and text overlay is specified at the level of detail you'd find in a professional shot list.

Step 3: Generation and Output

You see a full preview of the production prompts before anything generates. Timestamped scenes are displayed with element references highlighted, so you know exactly what each video will contain.

Choose your aspect ratio — 9:16 vertical for TikTok and Instagram Reels, 16:9 landscape for YouTube, 1:1 square for feed posts — toggle standard or pro quality, and optionally enable ambient audio or TTS narration. Then hit generate.

Dobidy sends all three prompts and your product images to Kling O3 in parallel. Your photos are passed as structured reference elements, not pasted onto frames. The model understands the shape, color, and proportions of your product and integrates it into each generated scene naturally.

Generation takes a few minutes. You can close the tab — Dobidy tracks all three jobs and picks up where you left off when you return.

The output is three 10-second video ads — one per scenario — each featuring your real product in a directed scene with intentional camera movement, lighting, and composition. No stock footage. No templates. No timeline editing. Three different creative angles to test, from a single session.

What This Replaces

To put the workflow in context: producing a comparable 10-second product video ad traditionally involves writing a brief, hiring a videographer or motion designer, scheduling a shoot or sourcing stock footage, editing, revisions, and export. That process typically costs $500-2,000 and takes 3-5 business days.

With Dobidy's Omni Tool, the active work is under 10 minutes — a two-minute chat, a photo upload, a quick review of three scenarios, and a button press. The AI handles scriptwriting, shot planning, camera direction, and video rendering — producing three finished ads in one go.

The creative strategy is still yours. You decide the audience, the platform, the tone, and which ads to run. The AI handles the production craft that used to require a team.

Beyond Ad Creation: More Tools, One Credit System

The Omni Tool workflow described above is Dobidy's flagship, but it is not the only tool on the platform. Spatial Story turns a series of product or space photos into walkthrough-style videos — useful for real estate tours, interior showcases, and immersive product presentations. An AI Avatar tool for spokesperson-style videos is also on the roadmap.

All tools share a unified credit system. The Omni Tool costs 300 credits for a batch of 3 video ads. Spatial Story and Time Story start free with a basic model (0 credits, 480p with watermark) and offer higher-quality models at 20-50 credits per clip. Paid plans start at Basic ($9/month, 400 credits) and Pro ($45/month, 2,000 credits) — so you can use whichever tools fit your needs without managing separate budgets.