World's First AI Risk Manager

• 

  Homepage with prompting hints and templates

• 

  Case studies and agents under the hood

• 

  Professional risk management advice

• 

  Benchmarking against public LLMs

• 

  Alignment with global standards and regulations

Inspiration

The inspiration for SIMON was born from an award winning Australian risk manager escaping from war. Our founder resigned from leading risk management and insurance for a $20 billion Russian owned corporation with a large team to a small Swiss, agile company focused on fighting climate change. In this new environment, resources were scarce. He could no longer afford the deep risk analysis for insurance, contracts, and new projects that he once took for granted. This wasn't just a downgrade; it was a barrier to making sound, secure decisions for a mission-critical business.

He turned to AI when GenAI first came out. The promise was immense, but the reality was disappointing. Public LLMs, while broadly capable, were notoriously bad at risk management. They offered generic, "popular" advice that was often impractical or outright wrong when it came to finding risks, making risky decisions and buying insurance. This was the "aha!" moment: a clear need for specialized, expert-level AI that understood the nuances of risk better. He first developed a chatbot, trained on a personal library of over 1,000 articles and 1,500 videos. This increased new risk team productivity 3X. He then developed a suite of task-specific AI risk management tools. The time invested in building them was repaid tenfold in hours saved and value created.

This mission resonated deeply within the professional risk community. Tapping into a network of 15,000+ risk managers from the annual RISK AWARENESS WEEK, which the founder created in 2019, a passionate team of senior risk professionals and developers came together. We saw this hackathon as the perfect catalyst to unite our separate, proven AI risk agents - which had already generated over $5 million in savings for companies - into a single, intelligent platform.

Our ultimate inspiration is twofold: to supercharge existing risk teams, and to democratize risk management. We want to make the expertise of a 10-person risk department, including quantitative PhDs, accessible to any founder, company, or government agency that needs to protect its mission, buy better and cheaper insurance, make better decisions under uncertainty and create value with confidence.

What it does

SIMON is not a chatbot; it is a full-stack AI risk manager powered by a team of over 20 specialized agents, built to save your company millions. When a user describes a situation - a new project, a complex contract, or a strategic decision - SIMON’s intelligent coordinator instantly understands the request, assembles the right team of AI agents, and executes a complex workflow to deliver a comprehensive analysis in minutes, not weeks.

This system goes far beyond the capabilities of public LLMs, SIMON’s agents perform highly specific, expert-level tasks. For example:

• Risk Radar: Functions as an instant, virtual stakeholder sounding board. It systematically analyzes any business scenario from multiple perspectives to uncover non-obvious risks you would otherwise miss, delivering in minutes what would take a human team days to achieve. Imagine, an honest, private, only for your eyes Board of directors, that see risks you don't see.
• Mitigation Maker: Turns risks into an actionable plan. It performs a deep root cause analysis and generates practical, cost-effective mitigation strategies based on industry-specific data and real-world case studies.
• Governance Builder: Creates a board-ready, fully customized risk management policy in seconds. It aligns with global standards like ISO 31000 or COSO ERM and weaves your company’s specific mission and context into a document that is ready for immediate implementation.
• Risk Translator: Ensures critical insights are understood by everyone, from engineers to the board. It transforms technical risk data into clear, persuasive, actionable intelligence tailored to the psychology and motivations of any corporate audience.

These are just four of the 20+ agents that work in concert to identify threats, quantify impacts, create scenarios, and even write a risk management joke in the voice of your favorite comedian. SIMON is a co-pilot that performs at the level of a world-class, 10-person risk team, delivering hard-cash savings and a true strategic advantage.

In risk management generic AI is not just insufficient; it's dangerous. That's why we moved beyond anecdotal evidence and developed a rigorous, objective framework to evaluate SIMON's performance against other models, including ChatGPT 4.1. We believe that AI tools without proper, transparent evaluation are a liability.

Our methodology scores AI responses against five core principles of effective risk management:

1. Decision Integration: Does the analysis connect directly to business decisions?
2. Practical Actionability: Are the recommendations concrete and implementable?
3. Quantitative Rigor: Is the mathematical and statistical approach sound?
4. Avoids RM1 Pitfalls: Does it reject useless compliance rituals in favor of actual risk reduction?
5. Business Context: Does it understand real-world organizational constraints and priorities?

In head-to-head benchrking, SIMON achieved an overall score of 9.5/10 compared to ChatGPT 4.1's 6.6/10. Read full benchmarking study

How we built it

To build a complex app like SIMON this fast, we needed a secret weapon. Our key was Bolt.new, a development platform that automatically generated 95% of our application's code. This let us start with a high-quality foundation and skip months of normal coding work.

We chose modern and reliable tools to build a scalable, launch-ready application:

• Our fast user interface is built with React, TypeScript, and Vite.
• For our entire backend, we relied on Supabase. It gave us our PostgreSQL database, secure logins, and serverless Edge Functions. This powerful foundation makes us a strong candidate for the Startup Challenge: Use Supabase.
• The frontend is deployed with Netlify, enabling continuous, automated deployment from Git and qualifying us for the Deploy Challenge: Use Netlify.
• To prepare for launch, we also integrated the RevenueCat mobile SDK to handle future subscriptions (Make More Money Challenge) and used Entri to set up our custom domain (Custom Domain Challenge).

This strategy of code automation was critical. It allowed us to pour all our energy into the most important part of the project: designing, building, and coordinating the 20+ AI agents. We believe this multi-agent system represents a Creative Use of AI that goes far beyond simple LLM wrappers.

Challenges we ran into

Our challenges were threefold. Technically, we had to perfect our security and speed. We dedicated significant time to hardening our Supabase Row Level Security to guarantee absolute data privacy for our users. We also worked relentlessly on reducing AI agent latency, optimizing our 20+ agent workflows to ensure complex insights were delivered in seconds, not minutes.

The final, and greatest, challenge was the clock. As a team of passionate professionals dedicating our nights and weekends to this project, we faced a feverish race against time. We were constantly incorporating a flood of valuable feedback from our risk community, iterating on the product right up until the final moments to build a tool they genuinely need.

Accomplishments that we're proud of

Above all, we are proud that SIMON is not just an idea, but a proven solution already making a real-world impact. Our proudest accomplishments are measured in user success and industry recognition.

First, the core AI agents that power SIMON have been used over 17,000 times by more than 1,000 users across the globe. This isn't just about usage; it's about tangible value. Our users have reported saving a collective $5M in insurance costs - not by cutting corners, but by using our tools to find higher quality and broader insurance coverage at a lower price.

Second, this real-world success was formally recognized by the industry. We are incredibly proud that one of our core agents was named a finalist in the 2024 European Risk Management Awards, validating it as a top innovation in the field.

But our greatest accomplishment is a human one. We are most proud of empowering and upskilling risk professionals. We’ve built SIMON to give managers - especially those in complex regulatory environments with limited resources - the analytical power they need to excel and add more value. That is the ultimate measure of our success.

What we learned

Our most profound learning during this hackathon was a stark realization of the low maturity of AI adoption within the risk and insurance profession.

We learned that building a powerful tool is only half the battle. There is a clear and urgent need for greater education to build trust and show professionals how to integrate AI into their workflows effectively. This has solidified our mission: SIMON isn't just about providing a revolutionary product, but about leading the charge in upskilling an entire industry to embrace the future of risk management.

What's next for World's First AI Risk Manager

We have a clear, three-phase plan to scale SIMON from a revolutionary tool into the global standard for risk management.

Phase 1: Immediate Community Onboarding. Our immediate priority is to deliver SIMON to our existing, highly-engaged community. We will begin onboarding the 15,000+ members of RISK AWARENESS WEEK, gathering feedback and rapidly iterating to ensure SIMON solves real-world problems at scale from day one.

Phase 2: Enterprise Readiness. Next, we will focus on the core needs of enterprise clients. This involves two parallel streams: developing a secure, on-premise corporate version of SIMON that can be deployed on private servers, and fundamentally enhancing the "brain" of our system - the intelligent router - to execute increasingly complex workflows with even less human intervention .

Phase 3: The Autonomous Risk Manager. Our ultimate vision is to evolve SIMON beyond a reactive tool into a proactive, autonomous virtual cloud risk manager. In this future state, users will no longer interact through chat. Instead, they will create persistent "artifacts" - a project, a legal entity, a contract - and upload the relevant documents once. From that moment on, SIMON will take over, autonomously identifying risks, generating a full suite of compliance documents, and continuously updating them in real-time as conditions change. It will become a system that doesn't just answer questions, but manages risk on its own.

This vision to build the central nervous system for risk management and insurance is why we believe SIMON has a clear path to becoming a category-defining company. It's why we are confident that SIMON is not just a project, but a strong candidate for the Future Unicorn and Most Likely to Get Funded awards.

Built With

• apis
• bolt
• eslint
• framer-motion
• git
• lucide-react
• netlify
• postcss
• postgresql
• react
• react-router-dom
• recharts
• revenuecat
• supabase
• supabase-auth
• supabase-edge-functions
• tailwind-css
• typescript
• vite