London-Based Organisation Challenged 21 Teams to Build Mental Health Software in 72 Hours – Then Asked Engineers from Okta, Visa, and Sonos Whether Any of It Was Safe to Deploy

MINDCODE 2026, organised by East Ham-headquartered Community Interest Company Hackathon Raptors, produced passive burnout detectors, AI-coached wellness platforms, and ethical mental health companions – raising questions about what rapid software development means for an NHS under pressure and a workforce reporting record levels of stress

Britain’s mental health crisis is well documented. NHS waiting lists for psychological therapies exceeded 1.9 million referrals last year. Workplace absenteeism linked to stress, anxiety, and depression costs UK employers an estimated £56 billion annually. And the technology sector – which promises to solve these problems – continues to ship mental health applications that ask overwhelmed users to do the one thing they cannot: take action.

MINDCODE 2026, organised by Hackathon Raptors, a Community Interest Company registered in England and Wales (CIC No. 15557917) and headquartered in East Ham, London, asked whether a different approach was possible. The event ran from 27 February to 2 March 2026, challenging twenty-one teams across multiple countries to build software where mental wellness serves as the primary design constraint – not a feature bolted onto an existing product. Thirty-eight engineers, including professionals from Okta, Visa, ADP, Sonos, Allegis Group, Apple, Amazon Web Services, and Microsoft, then evaluated every submission for engineering quality, ethical soundness, and real-world applicability.

The results suggest that the most effective mental health tools are not the most feature-rich – they are the ones that require the least effort from users who are already struggling.

What the Teams Built

The scoring framework weighted Impact & Vision (35%), Execution (25%), Innovation (20%), User Experience (15%), and Presentation (5%). A total prize pool of $2,500 was distributed across placement awards, category excellence, and a community vote.

First place went to MindMirror by team Taurus at 4.312 out of 5.0, sweeping four of five category awards (Best Impact & Vision, Best Innovation, Best User Experience, and Best Presentation). The system monitors digital footprints passively through a browser extension integrated with GitHub and Google Calendar, identifying burnout patterns before users recognise them. When deterioration is detected, the system triggers interventions – dynamic monochrome mode that strips colour from the screen, anti-doomscroll blocking that interrupts compulsive browsing – without requiring the user to acknowledge they need help. Ishu Anand Jaiswal, one of the evaluators, described it as a “standout project” where “passive burnout detection via digital footprint is highly innovative.”

Second place went to MeFirst by team LoopTroop at 3.850, earning Best Execution. The platform combines human wellness coaching with AI-assisted support, featuring structured onboarding, coach matching workflows, and a values wheel approach for tracking progress. Its architecture was designed with healthcare data protection requirements in mind – a consideration that distinguishes it from the majority of consumer wellness applications currently available in the UK market. Karthick Cherladine praised the “HIPAA-aware design and session management features” as showing “good consideration for real-world deployment.”

Third place went to Serenity by Sanjay Sah at 3.831, an AI mental health companion that distinguished itself through ethical architecture: AES-256-GCM encryption for journal entries, crisis resource integration, accessibility compliance, and AI safety guardrails that prevent the system from offering clinical advice beyond its competence. Saugat Nayak assessed it as “a thoughtful and well-designed mental health companion that demonstrates a strong balance between technical implementation and ethical responsibility.”

The Approaches That Scored Highest

Four additional projects formed a competitive Excellence Tier. MindLint by Team Batman (3.800) applied what developers call “cognitive observability” – monitoring mental strain as a system metric, the same way engineers monitor server performance. Haven (3.762) built gamified wellness engagement. MindTrace (3.731) earned praise for exceptional user experience design. REFRAME by beTheNOOB (3.719) embedded wellness support directly into workplace communication tools – Slack, Discord, and support dashboards – intercepting high-stress messages before they are sent.

Ramprakash Kalapala called REFRAME “an outstanding execution of a powerful idea” that “lets you experience it in the exact contexts where it’s needed.” The approach reflects a broader pattern observed across the strongest submissions: rather than asking people to download yet another wellness application, the most effective tools meet users inside the software they already use for work.

What This Reveals About Mental Health Technology

Three patterns emerged consistently across the twenty-one submissions – patterns that carry implications beyond a hackathon context.

Passive detection outperformed self-reporting. MindMirror and Echo (a separate project that analyses typing patterns as biomarkers for cognitive strain) both built systems that identify burnout without requiring user action. The insight is structural: self-reporting fails precisely when users are most overwhelmed. For a health system that relies heavily on patients recognising and articulating their own symptoms, passive monitoring through behavioural signals represents a fundamentally different detection model.

Ethical architecture correlated with engineering quality. The projects that treated data protection, consent, and safety as first-class architectural concerns consistently produced stronger code and more credible deployment paths than those that treated ethics as a documentation exercise. In a regulatory environment shaped by the UK Data Protection Act 2018 and the ICO’s guidance on health data processing, this correlation suggests that compliance-first design may be a competitive advantage rather than a constraint.

Integration beats installation. The strongest submissions embedded wellness support into existing workplace tools rather than building standalone applications. For UK employers seeking to address the £56 billion cost of workplace mental health, this pattern suggests that the most effective interventions may not look like “mental health apps” at all – they may look like features inside the communication and productivity tools employees already use.

The gamification question remains open. Several submissions experimented with gamified approaches to mental wellness. The Room by Thrixel (3.500) earned praise as “one of the most visually refined submissions” with ML-ranked journaling cards. Rewordit by Team Dua (3.512) took a different approach – catching high-stress messages before sending across platforms including email and messaging services. Whether game mechanics help or trivialise mental health engagement divided evaluators, though projects grounding gamification in clinical frameworks scored consistently higher.

The Evaluation Panel

The evaluation panel included 38 engineers organised across three scoring batches. Senior evaluators included Arun Kumar Elengovan (Director of Security Engineering, Okta), Madhushree Kumari (Visa), Nikita Klimov (ADP), Nandagopal Seshagiri (Okta), Dinesh Kumar Garg, Siarhei Krupenich (Allegis Group), and Manushi Sheth (Sonos), supported by evaluators including Sarthak Shah, Milind Jagre (Apple), Suprakash Dutta (Amazon Web Services), Giridhar Raj Singh Chowhan (Microsoft), and others from across the global engineering community.

About Hackathon Raptors

Hackathon Raptors operates as a UK Community Interest Company (CIC #15557917) headquartered in East Ham, London. The organisation conducts 25+ international hackathons annually, attracting participants from 85+ countries. Its fellowship model brings together engineers from leading technology companies as evaluators, ensuring that competitive outcomes reflect expert assessment of both engineering quality and real-world applicability. MINDCODE 2026 challenged teams to build software where mental health and human wellbeing serve as the primary design constraint. Further information is available at raptors.dev.