We spent years building image generation pipelines the hard way. So we built the API we wished we had.
We spent years stitching together image-generation pipelines. So we built the API we wished we had.
Hey Indie Hackers 👋
We're Dharmendra and Dharmik, two brothers and engineers who have been building software together since 2016.
Over the years, we kept running into the same problem.
We'd need to generate images programmatically - social cards, reports, product images, OG images, certificates, creator stats, etc.
The basic template part was easy.
Then someone would say:
"Can we put a chart in it?"
And suddenly we were combining a template engine, charting library, headless browser, rendering code, queues, storage and a bunch of glue code.
We kept thinking: Why isn't there one API that just does this?
So we built Imejis.io.
The basic workflow is: Design a template → send data → get an image.
You can create the design visually, then use the API to generate images from JSON/data.
But the part we're most excited about is that data visualisation is a first-class part of the rendering system.
Imejis currently supports 24+ component types, including:
- Bar, line, pie and area charts
- Candlestick charts
- Treemaps
- Sankey diagrams
- KPI cards
- Tables
- Maps
- Heatmaps
- Funnels
- Gantt charts
- Radar charts
- And more
So you can build things like personalised financial cards, creator statistics, leaderboards, analytics reports, social cards and other data-driven visuals without building your own rendering pipeline.
We've also added an MCP server.
That means AI agents can work with Imejis templates and generate finished images as part of an agent workflow.
We're particularly interested in this because we don't think every visual problem should be solved with generative AI.
Sometimes you don't want the AI to "imagine" your chart.
You want:
same template + same data = same image.
Accurate numbers.
Consistent branding.
Predictable layouts.
Deterministic output.
We've been bootstrapping Imejis as just the two of us and have already had it running in production for a real platform customer.
Now we're preparing for our Product Hunt launch.
But before we launch, we'd love some honest feedback from other founders and developers here.
A few things we're trying to figure out:
- Is data visualization inside an image-generation API actually a meaningful differentiator?
- Would you use something like this instead of building your own Playwright/Puppeteer pipeline?
- Do you see MCP/AI agents becoming a meaningful distribution channel for developer infrastructure?
- What use case would make you pay for this?
Here's what we're building:
Imejis.io - programmatic image generation for templates, data visualisation and AI workflows.
We're sharing the journey here as we get closer to launch.
If this problem is relevant to what you're building, I'd love to hear how you're solving it today.
