Award-Winning AI Technologist
Building AI for Good · Social Enterprise · Human Centred Model Augmentation
I'm Charlie Garnish — an AI technologist, social enterprise founder, musician, and advocate for the LGBTQ+ community.
I build products for situations where ordinary software usually falls short: sensitive support journeys, community safety, identity, sex-positive commerce, neurodivergent access, and AI systems that need firm boundaries.
As the founder of Glynda CIC, I lead a community interest company behind multiple LGBTQ+ wellbeing platforms. That includes Ask Bobby, an AI-assisted support directory signposted to over 1 million active monthly users; DorothyAI, a sex-positive product discovery system for Friend of Dorothy; and Glynda.org, a private chemsex support platform backed by psychologists and harm reduction experts.
My Human Centred Model Augmentation (HCMA) approach keeps AI close to human knowledge, lived experience, accessibility, and safeguarding, so the technology stays useful without drifting into theatre.
I also advocate openly around chemsex and its impact on the LGBTQ+ community, contributing to education, harm reduction, and the wider conversation around queer health.
A portfolio of live and in-development platforms across LGBTQ+ wellbeing, social connection, creator verification, health, commerce, neurodiversity, and operations.
Run by Glynda CIC, a social enterprise dedicated to technology for good and supporting the LGBTQ+ community. Glynda.org is a private chemsex support and wellbeing platform for gay men navigating psychosexual and substance-use challenges, providing structured journaling, practical toolkits, and safe community spaces.
View Case Study
The UK's LGBTQ+ support and connection directory, signposted to over 1 million active monthly users. Built for You Are Loved to help people describe what they need in plain language and find vetted support.
View Case Study
A sex-positive conversational AI built for Friend of Dorothy, a gay men's supplement brand. DorothyAI helps customers ask candid questions about topping, bottoming, and sexual wellness, then turns that context into useful product discovery. Includes custom Shopify Liquid theme development.
View Case Study
ADHD coaching, personal support, and workplace neurodiversity training. Built with Georgia Franklin-Cook to help neurodivergent people navigate brain fog, overwhelm, executive functioning, and daily life without making the service feel clinical or confusing.
View Case Study
A nutritionist-led gut health and fertility specialist platform blending informed nutrition science with digital technology for personalised wellness.
View Case StudyA full technology rental operations suite covering CRM, order management, logistics, asset tracking, MDM, and IoT SIM automation. Built to reduce device prep time by up to 95%.
View Case Study
A decoupled verification layer that helps people and creators link images, profiles, and original work to a verified identity. Link-in-bio meets authenticity and takedown support.
View Case Study
A new Glynda CIC app for the gay and queer community, built around safer connection, social gravity, consent-aware interaction, and real-world belonging.
View Case StudyThe IT Awards
Recognised for Ask Bobby's reach to over 1 million active monthly users in the UK, advocacy around chemsex and community impact, and responsible AI work in LGBTQ+ wellbeing.
Ask Bobby signposted to over 1 million active monthly users across the UK
Open advocacy on chemsex, harm reduction, and queer health education
AI platforms driving real outcomes in LGBTQ+ wellbeing and support services
A framework for building AI systems that start with the human — not the data.
HCMA is my approach to responsible AI development. It places lived experience, accessibility, and community safety at the core of every model and system. Rather than retrofitting ethics into existing architectures, HCMA embeds human perspectives from the first line of code.
Models trained on and validated by the communities they serve — not just datasets about them.
AI that removes barriers rather than creating them. Every interaction designed for the widest possible audience.
Clear boundaries between human and machine intelligence. Users always know when AI is in play.
Proactive safety design that anticipates misuse, bias, and unintended consequences before deployment.
Articles, research, and perspectives on AI ethics, community technology, and LGBTQ+ advocacy.
On the emerging digital risks facing LGBTQ+ and neurodivergent communities — from low-quality automated content and hallucinating chatbots to platform accountability, safeguarding, and the hidden cost of poorly chosen technology.
The intersection of LGBTQ+ lived experience and AI development isn't a niche — it's a necessity. Here's why diversity in AI isn't just ethical, it's functional.
Moving beyond ethical AI checklists to a structural approach that embeds human perspective into model architecture from day one.
How an intelligent directory platform reached over 1 million users and what it taught us about building technology for marginalised communities.
Every stream directly funds LGBTQ+ support. For every five streams, one conversation with Glynda is paid for — real impact, powered by music.
Five Minute Fantasy, Everybody's Talking, and Labubu Crush are LGBTQ+ pride anthems — safe-for-work celebrations of queer joy and identity. There's also a sex-positive side — tracks like Good Boy, Feral, Backseat Bend, and No Type — unapologetic, provocative electronic bangers best enjoyed out and about or in the car.
I choose tools for speed, resilience, privacy, and maintainability. The stack matters, but the real skill is knowing what to use, what to avoid, and how to turn a messy idea into a platform people can trust.
I use Google Cloud when a product needs to be reliable without becoming over-engineered. It gives me managed infrastructure for APIs, scheduled work, analytics, AI services, and integrations, while keeping operational overhead low enough for social enterprises, startups, and specialist teams.
Firebase is often the fastest route from prototype to production, if it is designed properly from the start. I use it for authentication, real-time data, serverless workflows, hosting, storage, and access control, with particular care around permissions, multi-role products, and sensitive user journeys.
Most products only become useful when they connect cleanly to the rest of the organisation. I build payment flows, email journeys, booking systems, analytics, Shopify integrations, and operational dashboards so the platform does not sit in isolation from the work it is meant to support.
For mobile-first products, I favour cross-platform native development where one well-structured codebase can serve iOS and Android without losing the feel of a real app. This is the direction behind Meatspace and other products that rely on camera access, notifications, account flows, and fast iteration.
My AI work is built around controlled systems, not novelty demos. I design conversational tools, structured retrieval, safeguarding layers, and human review points so AI can support real users without pretending to know what it does not. That is the thinking behind Ask Bobby, DorothyAI, and the wider HCMA approach.
I build for people who may be sharing sensitive things: health, identity, sexuality, location, money, or support needs. Privacy has to be practical, visible, and designed into the system before anyone signs up.
My work often sits close to vulnerable moments. That changes the standard. I think about what data is collected, who can access it, how long it is needed, what happens if something goes wrong, and whether the product can deliver value without tracking people unnecessarily.
I treat secure access, device hygiene, MFA, least-privilege permissions, protected admin areas, and sensible deployment practice as normal product work. These are not glamorous features, but they are what make sensitive platforms credible.
I design with UK GDPR and the Data Protection Act in mind, but the aim is not to hide behind policy documents. The aim is to collect less, separate access clearly, explain what is happening, and make sure personal data is only used for the purpose the user expects.
Support platforms, dating products, health tools, and creator verification systems all carry different risks. I map those risks into the user journey: consent, moderation, audit trails, data retention, account recovery, reporting, and the edge cases most templates ignore.
I do not build support services that quietly turn distress into advertising data. When I need analytics, I prefer restrained measurement that answers product questions without following people around the internet.
If someone comes to a platform for help, their vulnerability should not become a marketing signal. That principle shapes how I approach every product in the Glynda CIC ecosystem.
I help organisations turn complex, sensitive, or ambitious ideas into products that work in the real world.
Conversational systems, retrieval workflows, safety layers, and AI features that are scoped around the actual user journey rather than bolted on for novelty.
Product architecture for platforms with users, roles, payments, admin tools, data rules, integrations, and the messy operational details that appear after launch.
Bespoke web apps, dashboards, portals, mobile-ready products, and internal tools built around the way your organisation actually works.
Accessible interfaces for people under real cognitive load: clearer flows, calmer content, better defaults, and support for neurodivergent users.
Practical AI governance, product audits, and system design for teams that need AI to be useful, explainable, and safe enough for sensitive contexts.
Technology strategy for CICs, charities, founders, and specialist teams who need impact, operations, and sustainability to work together.
Bring me complex product problems: sensitive user journeys, AI that needs guardrails, platforms that need to scale, or ideas that need turning into something people can actually use.
charlie@glynda.org
