IgniteAI on Google Cloud

Last updated: 18 September 2026

IgniteAI (igniteai.in) is a product of The Jaggery Point, Bengaluru, India — a small, bootstrapped company that also operates café outlets under the same name. IgniteAI is the software business; this page describes it for reviewers and partners in plain terms: the problem, the customer, the product, where it stands, and how it runs on Google Cloud.

The problem, and who has it

Direct-to-consumer brands live or die on short vertical video ads for Instagram Reels, YouTube Shorts and TikTok — and they need a steady stream of them. A single creator-style ad from an agency or a freelance videographer takes days and costs more than most small brands can spend per creative. So they post less, test less, and pay more per sale.

Our customers are founders and marketing leads at small and mid-sized D2C brands (food and beverage, skincare, jewellery, fitness, home, pet products), the marketing agencies that serve them, and course or workshop businesses that sell online. India first; English-speaking markets next.

Industry

Marketing technology — specifically generative media for e-commerce advertising. IgniteAI sits between the brand's product catalogue and the ad platforms: it turns one product photo into a finished, platform-ready video ad.

What the product does

A user uploads one product photo and a sentence about the product. IgniteAI then:

  • researches the product and writes a short creator-style script (Hook → Feature → Call to action);
  • casts one consistent AI creator and renders a storyboard — three scene stills the user can edit or re-roll for free before any video is generated;
  • animates each scene into video with a spoken line, adds karaoke captions and music, and delivers a 9:16 ad of 15–30 seconds in roughly five to ten minutes.

Around the core product sit four free tools that need no account — an Ad Hook Analyzer, “Roast My Ad”, a break-even ROAS calculator and an AI Product Photo Generator — and a public MCP server so AI assistants can use the tools on a user's behalf.

Screenshots

The IgniteAI storyboard editor: Hook, Feature and CTA scenes with scripts, re-render controls and the properties panel
The storyboard editor. Each scene shows its line and still; the user can re-render any scene before video is generated.
Frames from a finished 21-second IgniteAI ad, one every 2.5 seconds, with captions and end card
A finished ad, one frame every 2.5 seconds: the same creator across scenes, product in hand, captions, end card with the brand's website.
The free AI Product Photo Generator page on igniteai.in
The free AI Product Photo Generator — one of four public tools.

A real output video plays on the homepage (“Fig. 3 — The Proof”). The app itself is at app.igniteai.in; new accounts receive ten free credits.

Stage of development

Live, in early access. The storyboard-based pipeline has been in production since May 2026; the free tools since July; the product photo generator since September. Self-serve purchases are paused while we scale capacity, so new customers start on free credits and ask for more from inside the app; existing paying users keep their credits.

As of 18 September 2026, from our production database:

  • 29 registered accounts, with organic sign-ups arriving through the free tools — including one this week referred by ChatGPT;
  • 21 finished ads rendered through the storyboard pipeline, 32 storyboards created;
  • 23 leads captured by the free tools and waitlist; 14 product photos generated by the newest tool;
  • real cost of goods measured per run (about $2.30 for a 15-second ad), tracked server-side on every generation.

Team: one founder building and operating the product full-time, with AI coding agents. No outside equity funding.

How IgniteAI runs on Google Cloud

Everything below is in production today, not planned.

  • Cloud Run — the FastAPI backend (API, MCP server, and the video pipeline workers).
  • Cloud Tasks — durable, retry-safe dispatch of long-running storyboard and render jobs.
  • Firestore — users, credit ledger, storyboards, runs, funnel events, leads, rate-limit state.
  • Cloud Storage — generated stills, clips and finished ads; customer product photos kept privately behind short-lived signed URLs.
  • Firebase Authentication and Hosting — sign-in, the web app and this marketing site.
  • Secret Manager — every API key and token used by the backend.
  • Gemini — product research with Google Search grounding, script writing, visual-DNA extraction, vision quality checks, and image generation for storyboard stills and product photos.
  • Veo 3.1 — image-to-video for every scene, with native speech.
  • Vertex AI Search — a curated datastore of ad patterns that grounds the script writer.
  • Cloud Logging and Cloud Build — observability and deployments.

Video generation is the dominant cost and the reason capacity, not demand, is our current constraint: Veo quota on the Gemini API is the limit on how many ads we can render per day.

Contact

Nikhil Patil, founder — business@igniteai.in. Company website: thejaggerypoint.com. Product: igniteai.in.