Editorial Policy

Who reviews our content

Every social story, blog article, and product page on SocialBlossom is reviewed by Rajini Darugupally, M.Sc. SLP (AIISH), a practising speech-language pathologist with 11+ years of clinical experience. She also practises at Wellness Hub, where she has been since 2017.

Rajini is the only named clinical reviewer on this site. We do not claim a larger editorial board than we have.

How content is created

Social stories (products)

Each social story is generated by an AI pipeline that takes a topic, age range, and skill focus as input, drafts the page-by-page text and accompanying illustration prompts, and produces images using Google’s Gemini image model. The story metadata, page text, and activity pages are then reviewed by Rajini before the product is marked published.

Blog articles

Articles are drafted by an in-house pipeline that uses OpenAI’s GPT models, structured against a fixed template (TL;DR, sections, activities, “Try It This Week” plan, FAQs, sources). The draft is run through a lint check (clinical-language flags, word-count targets, required widgets), then reviewed by Rajini and marked “reviewed” with a published date and last-reviewed date you can see on every article.

Product page descriptive copy

The “How to use this story” section on each product page is auto-generated from the underlying story content (sentences, activity types, skill tags) using GPT-4o-mini, against a prompt that requires strengths-based, neurodiversity-affirming language and the specific story’s context. The output is reviewed alongside the story itself.

How we use AI — and how we don’t

We use AI to draft. We don’t use AI to publish. A real human (Rajini, plus Akash for engineering checks) reviews everything before it’s live, and the final responsibility for what goes on this site is ours.

Where AI is used

• Drafting social story page-by-page text
• Drafting blog articles against a fixed structural template
• Generating illustrations (Gemini image model)
• Generating the “How to use this story” descriptive block on each product page
• Generating the educator / parent / OT “perspective” cards on each product page (see next section)

Where AI is NOT used

• Rajini’s named clinical reviews and bylines (those are written or approved by Rajini in her own voice)
• Customer testimonials (user-submitted only; never invented)
• Pricing, refund decisions, or any clinical recommendation about a specific child

About the “perspective” cards on product pages

On each product page you’ll see Rajini’s named clinical review alongside three additional cards labelled Special Educator perspective, Parent perspective, and Occupational Therapist perspective.

These are not reviews from named real people. They are composite perspective summaries written by GPT-4o-mini using the specific story’s content and our editorial prompt — designed to give you a quick read on how that story tends to land in three common use-cases (a classroom, at home, a therapy session). They are intentionally anonymous because the underlying viewpoints are general, not the testimony of a specific individual.

We chose this approach because we wanted multiple lenses on each story (clinical, classroom, home, OT) without inventing fake reviewers. If you ever see a real-name review on a product page, it is either Rajini’s or a customer-submitted review.

Sources we draw on

When we make claims about social-emotional skills, child development, or how social stories work, we draw on:

• The CASEL 5 framework for social-emotional learning (self-awareness, self-management, social awareness, relationship skills, responsible decision-making) — CASEL
• The social-stories tradition, originally developed by Carol Gray in the 1990s
• Peer-reviewed research on narrative-based social-emotional learning, speech-language therapy, and neurodiversity-affirming practice
• Rajini’s 11+ years of clinical practice

We do not fabricate citations. If a blog article cites a study, we link to its source. If we paraphrase a clinical principle, we name the framework or author behind it.

How we update content

Each blog article displays both its first-published date and its last-reviewed date. When evidence changes, when an article’s framing no longer matches our current position, or when a reader flags an error, we update the article in place and refresh the last-reviewed date.

Social stories are versioned. If we improve a story (better illustrations, clearer text, new activity), we ship the new version and existing buyers keep access to the updated PDF.

Corrections and feedback

If you spot an error, a clinical inaccuracy, an outdated reference, or anything that doesn’t feel neurodiversity-affirming, please tell us. We’ll review and either correct it, update the last-reviewed date, or — if we disagree — explain why and add a note on the page.

Email: sakash674@gmail.com

What this site is, and isn’t

SocialBlossom is an educational resource. It is not a substitute for individualised therapy, a clinical diagnosis, or medical advice. The activities and stories on this site support learning; they do not replace a professional’s assessment of your specific child.

See our full disclaimer for details.