.avif)
.svg)
Forum's Venture Studio brings together ambitious founders, brilliant ideas, and capital to build the best B2B SaaS businesses in the world, from 0 to 1. In addition to capital and an idea, we support founders with a full design and development team, go-to-market support, strategy, and back-office (accounting, legal, HR) support. We also offer full fundraising support, including narrative and pitch deck creation, investor outreach plans, and investor introductions. Our model is designed to help you move faster, develop better insights, and build companies that have a higher success rate than startups built in any other way.
At Forum Ventures’ AI Studio, every week we’re identifying and evaluating venture-scale ideas that align with major industry shifts and emerging investment trends, that we’re genuinely excited to build.
Lately we’ve been obsessed with the ability of the new models ability to automate complex, regulated workflows.
Until recently, AI was mostly used to automate simple, repetitive tasks, customer support chatbots, invoice classification, and basic data entry. But with the rise of more advanced models capable of reasoning across unstructured data, orchestrating multi-step processes, and learning from edge cases, we can now tackle complex, compliance-heavy workflows.
The obvious wins in AI have mostly been taken. The real opportunity now lies in tackling the non-obvious stuff. Complex, regulated, and error-prone workflows that legacy systems and human processes still struggle to handle.
Compliance Copilot for FedRAMP & StateRAMP
Context
FedRAMP and StateRAMP are government programs that certify the security of cloud products and services, which companies must obtain before they can sell into government. However, achieving certification is a long and expensive path, often taking 12–18 months and costing upwards of $2 million per application.
With the recent rescission of the Emerging Technology Prioritization Framework, early-stage vendors can no longer fast-track approvals. Meanwhile, agencies and vendors struggle with having to constantly monitor systems, get separate approvals for each individual application, and repeat the same processes across different cloud platforms.
The Market
The FedRAMP compliance market was valued at $3B in 2023 and is projected to exceed $6.3B by 2031. The number of authorized offerings has grown significantly. Regulatory tailwinds, including the 2022 FedRAMP Authorization Act Regulatory that strengthened the program and new directives mandating automation and machine-readable documentation, is pushing both agencies and vendors toward smarter, more automated tools.
The Opportunity
Our current thesis is to build a FedRAMP/StateRAMP automation platform tailored specifically for startups and mid-market SaaS companies. This platform will streamline the path to public sector compliance by accelerating approval timelines, automating documentation, and enabling real-time security monitoring across multiple cloud environments.
It will include features like automated generation of compliance documents, tools for working with third-party security assessors, and integrations for running security scans. The platform will also help vendors connect with potential agency sponsors, and provide real-time dashboards to track vulnerabilities and stay audit-ready. What’s normally a slow, manual process becomes faster, and more predictable
Why We’re Excited
As the current administration continues to prioritize partnerships with technology companies and the automation of workflows within government agencies, and as cybersecurity mandates intensify, the demand for faster, more cost-effective compliance solutions will only grow. Agencies are now required to accept automated compliance artifacts, creating a unique opportunity to define the modern infrastructure layer for cloud security compliance.
AI-Powered Duty Drawback Automation
Context
Duty drawback processes offer significant financial benefits, allowing businesses to reclaim up to 99% of duties, taxes, and fees paid on imported goods that are subsequently exported or destroyed. However, only a fraction of eligible duties are ever refunded due to the highly complex, manual nature of the process, which requires meticulous matching of import and export records, precise document preparation, and strict compliance verification. Recent geopolitical shifts, tariffs, and evolving regulatory frameworks have further increased the complexity and urgency of effective drawback management.
The Market
The Duty Drawback Service Market was valued at $21.3 billion in 2023 and is projected to reach $41.1 billion by 2031. Approximately 14,454 active licensed customs brokers operate in the U.S. - many of whom still rely heavily on manual, labor-intensive processes. This market has historically been dominated by consultants, whose comprehensive offerings don’t fit the nuanced needs of the middle market.
The Opportunity
An AI-powered duty drawback platform designed specifically to simplify and automate drawback claims, with a focus on substitution, unused and manufacturing drawback types. The platform will identify eligible claims through automated trade data analysis, effortlessly prepare and submit the necessary documentation for CBP compliance, and provide real-time tracking and predictive analytics to accelerate and claim approvals.
The vision is to build a fully automated system that makes duty drawback claims fast, easy, and accurate, enabling mid-market importers and brokers to secure refunds more quickly while maintaining strict regulatory compliance.
Why We're Excited
Growing tariff complexities and stricter regulatory enforcement have created a need for a simplified, automated duty drawback solution. Duty drawbacks represent a powerful but often underutilized source of "found money" for businesses. By dramatically reducing the time, complexity, and effort required to secure refunds, our platform taps into a clear and compelling value proposition: businesses will naturally adopt a solution that unlocks cash they are already entitled to with minimal friction. If we can streamline and accelerate the claims process, demand will follow.
Proactive Chargeback Prevention
Context
Chargebacks are no longer just a financial nuisance, they are becoming a significant compliance and operational risk, especially in regulated sectors like healthcare and financial services. Traditional tools rely on manual post-dispute processes that are slow, error-prone, and ill-suited to meet the demands of today’s complex regulatory and transactional environments. As regulatory scrutiny increases and first-party fraud grows, the limitations of reactive systems are more evident than ever. Industries with complex billing systems and strict compliance requirements are particularly exposed, creating a need for infrastructure that can identify and prevent chargebacks before they happen.
The Market
Global chargeback volume is projected to reach 324 million transactions by 2028, growing at an annual rate of 7%. In healthcare, chargebacks are typically triggered by billing errors, patient dissatisfaction, or insurance disputes, while for financial services face significant costs due to high transaction volumes, and complex dispute resolution processes.
While well-funded companies like Forter, Signifyd, Sift, Sardine, and Justt have gained traction, most existing solutions remain reactive, focused on resolving chargebacks after they occur. Even with the rise of AI in fraud and dispute management, the market still lacks truly proactive tools designed to prevent chargebacks before they happen.
The Opportunity
We're building a chargeback prevention platform tailored for regulated industries like financial services and healthcare, where chargebacks aren’t just about lost revenue - they come with serious compliance and operational risks. The platform will analyze transaction data and customer behavior in real time to spot and stop high-risk activity before it becomes a dispute. It’s designed to evolve with new fraud patterns and plug into existing systems smoothly, without disrupting compliance workflows or day-to-day operations.
Why We're Excited
Chargebacks are rising fast, and while retail brands have started adopting new tools, regulated industries are still seeing more incidents and more risk. In these compliance-heavy environments, the need for proactive prevention is urgent.
Synthetic Data for Asset-Based Lending Underwriting
Context
Asset-Based Lending (ABL) relies on accurate valuation of physical collateral, such as machinery, equipment, and inventory, but today’s methods are largely manual, expensive and prone to human error. The process typically involves in-person appraisals and field audits that can take days to complete and cost thousands per assessment. As a result, deal cycles are delayed, access to credit is limited, and operational burdens fall heavily on lenders, especially small and mid-sized institutions. Despite rising demand for real-time insights, the infrastructure for valuing and monitoring tangible assets remains fundamentally outdated.
The Market
The current Asset-Based Lending (ABL) market is valued at over $682 billion and is projected to grow to $1.28 trillion by 2033, fueled by increasing adoption among small and mid-sized enterprises, as well as rapid expansion in the manufacturing and logistics sectors. Banks, responsible for more than 40% of all ABL loans, are actively seeking better technology to manage collateral valuation risk.
The Opportunity
While ABL remains stuck in outdated processes, the synthetic data space is gaining traction, with startups like MostlyAI, Aindo, and YData raising capital for broad financial applications. Yet no one is addressing the unique needs of the ABL asset class, leaving a clear gap for disruption.
We’re building a synthetic data platform purpose-built for ABL underwriting and real-time asset monitoring. The platform uses computer vision to analyze images and videos of equipment for condition assessment and inventory verification, and leverages diffusion models to generate synthetic data based on historical appraisals, simulating asset performance over time. It also interprets unstructured data like valuation reports and liquidation records to support underwriting decisions.
Why We’re Excited
As lending expands to underserved sectors and economic uncertainty increases, lenders need faster, cheaper, and more accurate ways to manage collateral risk. Synthetic data offers not just efficiency, but a bridge to predictive credit modeling, automated audits, and even embedded credit insurance.
Ready to build?
We're seeking exceptional founders to tackle these four concepts head-on. With the support of a dedicated team and $250K in initial funding, you’ll have everything you need to transform these concepts into thriving businesses.
If you’re interested in building with Forum please apply through the corresponding founder-in-residence description. If none of these areas fit what you’re working on, feel free to apply through the Forum Studio Founder posting. Please highlight relevant experience and the area you’re most interested in.
While we’re actively building in these spaces, Forum Ventures is always open to hearing from founders with bold, original ideas. If you're working on an early-stage B2B SaaS company, even at the idea stage, we’d love to hear your pitch through the Forum Studio Founder posting.
Founder-In-Residence Postings
.jpg)
.jpg)
.avif)
