fintechforensics.com

What It Is

Fintech Forensics is not a self-service AI tool platform, but a boutique consulting firm based in Kigali and delivered end-to-end by its two founders. It positions itself at the intersection of forensic finance, risk control, sustainable finance, and production-grade AI, serving banks, fintech companies, and multinational enterprises. The website explicitly emphasizes “A practice, not a platform,” making it a better fit for organizations that need customized investigations, model deployment, and expert judgment rather than teams looking for ready-to-use SaaS.

Core AI Capabilities

Its AI capabilities cover the multimodal data needed for financial investigations: NLP/RAG can be used to parse contracts, emails, and chat records; ASR and speaker diarization can process calls and interviews; OCR and document AI can extract tables from invoices, statements, and contracts while detecting tampering, duplication, and inconsistencies. It also includes entity graphs, case memory, multi-agent investigations, hypothesis reasoning, and explainable outputs. The site emphasizes that models are built for specific problems rather than simply applying general-purpose LLMs, and that they can run behind a client’s firewall, in-country, or in offline environments.

Typical Use Cases and Delivery Model

Typical use cases include internal fraud investigations, detecting procurement and supplier collusion, identifying duplicate or ghost-supplier invoices, AML transaction tracing, ERP process optimization, automated risk anomaly detection, evidence package generation, as well as Fractional CFO and ESG finance consulting. In terms of process, the initial 30-minute call is free; in the first week, the team can ingest sample ledger or ERP data and deliver a written diagnostic within five business days; from weeks 2 to 6, it builds and deploys forensic AI or automation systems; ongoing collaboration can then continue on a Retainer basis.

Pricing and Availability

Public information only states that engagement models include project, retainer, and fractional hours, with no specific pricing disclosed. For enterprise clients, this model makes it easier to tailor work to case complexity; however, for teams with limited budgets or those hoping to quickly procure standardized software, the lack of price transparency is a drawback. Payment methods are not disclosed.

Pros and Cons

Its strengths lie in its deep focus on financial forensics, direct founder involvement, and emphasis on evidence-grade reporting, source traceability, audit trails, and regulatory usability. It also supports on-prem, in-country, offline, and lightweight GPU deployments, making it relatively friendly to sensitive financial data. The limitations are also clear: it is not a platform product, and delivery depends on consulting capability; the small team size may limit parallel projects; while the site mentions representative results such as “up to 95% anomaly detection accuracy” and “50% reduction in false positives,” these are mostly anonymized representative cases and lack public third-party benchmarks.

Who It’s For and Access from China

It is suitable for financial institutions and large enterprises with complex ledger, ERP, invoice, AML, or internal fraud investigation needs, especially those willing to implement customized AI solutions. It is less suitable for users who simply want to buy a general-purpose chatbot or low-cost automation tool. The main content does not specify access from China, network connectivity, or cross-border payment availability, so these remain unknown. For deploying similar capabilities in China, it may also be worth evaluating local risk-control AI vendors, audit consulting teams, or alternatives such as Palantir, Feedzai, ThetaRay, and SAS.