Ashish Sheth's Blog

I have been playing around with various image generation models for the last 2 years.  I used to spend lot of time to come up with a prompt which gets me the desired result  I learnt a lot about how to create effective prompts to get desired images after lots of trial and error, reading through official guides and learning from prompts that people share online. Finally automated the prompt creation by creating a tool which can easily generate prompts for all the tools that I was using myself. It's a Google Sheet. It provides dropdown menus for every parameter that matters such as art styles, camera angle, special effects etc. Pick your model, describe your image, and the prompt writes itself, structured exactly how that specific model wants it. The nine models it supports are: Midjourney, GPT Image 1.5, DALL-E 3, Stable Diffusion 3.5, FLUX.2 , Nano Banana Pro, Ideogram 3.0, Leonardo Phoenix 1.0, Seedream 5.0 It also supports following art styles: 3D Render / Pixar Style, Abstr...

Let's see five key patterns in which LLM Workflows are implemented. Prompt Chaining This workflow is used when we can break down a complex task into a sequence of steps, and each step can be processed as individual LLM call using the output of the previous step. Example: Multistep reasoning tasks such as summarize text, critique and then rewrite Code generation pipeline such as generate plan, write code, test code, explain output Router This workflow is used in cases where an agent classifies the intent/task type/domain of the input query and then delegates the request to another agent, tool or workflow.  Example: Multimodal agents such as those with separate text processing and image processing LLMs Multidomain assistants such as processing questions related to legal vs finance vs medical domains Parallelization The workflow pattern is useful when a task can be broken down into subtasks which can be handled concurrently by multiple LLM or agents. The outputs of these multiple call...