Finthesis AI: Why your financial reports arrive too late
- 3 hours ago
- 8 min read
A manager rarely lacks data. However, they often lack the ability to interpret it clearly at the right time. The income statement exists. The balance sheet exists. Exports exist. Tables exist too.
However, in many growing SMEs, financial information arrives too late, too raw, and with too little commentary. It serves to explain what has already happened, rarely to decide what needs to be done now or in the future.
This is precisely where Finthesis becomes interesting. The platform was already positioned in financial reporting, data visualization, dashboards, forecasting, consolidation, valuation and social or ESG indicators.
Its AI layer adds a more operational dimension : transforming financial data into written analyses, scenarios, recommendations, business plans and sector comparisons.
Finthesis also features an MCP integration with Claude to interact with financial data in natural language, as well as a connection with Perplexity to obtain benchmarks and sector analyses.
For an SME with a turnover of more than 500,000 euros, especially when it is growing quickly or developing internationally , the challenge is clear: to reduce the gap between accounting data and decision-making.
The real problem: financial data is lying dormant for too long.
In a well-structured company, accounting is not just about producing tax returns. It should help to understand profit margins, anticipate cash flow, measure the impact of recruitment, challenge a budget, compare scenarios, or prepare for discussions with a bank.
The problem rarely stems from a complete lack of tools. Many companies already use Pennylane , QuickBooks, Stripe, Shopify, Silae, Qonto, Dext, or other solutions. The workflows are in place. The data flows. But between the raw data and the decision, there's still too often a layer of manual analysis.
This is where Finthesis provides a relevant solution. The platform retrieves accounting data via API or import, structures it, and then transforms it into actionable tables, graphs, and financial indicators. Finthesis specifically mentions its ability to connect to accounting tools such as Pennylane , Sage, Odoo, ACD, Fulll, Inqom, My Unisoft, Tiime, and Cegid Loop.
The connection with Pennylane is particularly well-suited for Blendy clients. Pennylane centralizes accounting and financial flows. Finthesis then provides a more advanced layer for analysis, reporting, and forecasting.
The Finthesis documentation specifies that importing via API with Pennylane follows a simple logic: creation of the project, choice of the Pennylane API, connection, selection of the exercises to retrieve, then manual or automatic, monthly or weekly update.
READ ALSO
What AI really changes in Finthesis
We must avoid the classic trap: talking about AI as a gadget. In Finthesis , its interest lies mainly in three concrete uses.
1- The first concerns the analysis .
Finthesis states that its AI can detect anomalies, trends and points of concern in financial data.
For a manager, this function is valuable when it brings to light more quickly a margin drift, a cash flow strain or a variance with the budget.
2- The second concerns the annotated report.
Finthesis can generate annotated and customized reports, ready to be shared. This is a significant change for SMEs that produce spreadsheets without always being able to clearly articulate their analysis. A graph alone is not enough.
What matters is the sentence that explains why the margin is falling, why the working capital requirement is tightening, or why the result is improving despite stable sales.
3- The third concerns the projection.
Finthesis emphasizes forecasts, recommendations, business plans in natural language, and scenarios. The value lies not in the automated production of a document.
It is capable of quickly testing several hypotheses: recruitment, opening a new market, price increases, margin variation, investment, debt, fundraising or temporary decrease in activity.
The most advanced feature: querying your financial data in natural language
The most interesting new feature isn't the automatically generated report. It's the access to data through conversation.
Finthesis highlights its MCP integration with Claude . MCP, introduced by Anthropic , is an open standard designed to create secure connections between data sources and AI tools. In Finthesis 's case, this logic is used to query financial data in natural language, create scenarios, test forecasts, and challenge assumptions without leaving Claude.
For a manager, this changes the way they work. Instead of asking their consulting firm or CFO for a new spreadsheet for every question, they can make more direct requests:
“What happens if my gross margin drops by two points for six months?”
“What level of revenue do I need to reach to finance two sales hires?”
“Which items explain the difference between my budget and the actual amount?”
“What will the cash flow trajectory be if my customer payment terms are extended by fifteen days?”
This type of usage obviously does not replace financial analysis. It speeds up the initial reading and helps the right questions emerge more quickly .
The arbitration itself remains human. Finthesis makes this clear: AI assists, advises and automates, but the user retains control over the analyses and decisions .
Why Perplexity also changes the sectoral reading
The other point to look at closely concerns Perplexity . Finthesis presents a connection that is used to produce sector comparisons with market trends, margins , ratios , multiple recommendations and valuation elements.
This is particularly useful for SMEs that lack external benchmarks. Many managers know if their revenue is increasing. They sometimes know if their margin is declining. But they have more difficulty answering a simple question: am I underperforming compared to my market?
For an IT services company, a SaaS provider, an eCommerce business, or a growing SME, industry benchmarking adds context to internal data. A declining margin doesn't have the same significance if the entire sector is experiencing the same pressure, or if the company is the only one lagging behind. Valuation is also not determined in a vacuum. It depends on industry multiples, profitability levels, recurring revenue, and the accuracy of forecasts.
Finthesis does not automatically transform a benchmark into a strategy. But it shortens the path between data, comparison, and substantive discussion.
The tangible benefits for a growing SME
For an SME that is starting to exceed the artisanal management threshold, Finthesis responds to a very common tension: the manager wants information that is more readable, faster, more decision-oriented, without immediately recruiting a full financial department.
The tool becomes useful when the company needs to monitor its cash flow, structure a budget, prepare a business plan, produce investor reports, manage margins by activity, compare several scenarios, or present its figures to a management committee...
Finthesis also offers a very tangible benefit in terms of format. The platform indicates that reports can be exported as PDFs, Word documents, PowerPoint presentations, or shared directly through the platform. For a company supported by Blendy , this facilitates the production of readable materials for steering committee meetings, banks, investors, or partners.
The most significant gain lies elsewhere: improved financial discipline . When figures are clear, explained, and regularly updated, decisions cease to rely solely on the leader's intuition. Intuition remains valuable, but it becomes more robust when supported by quantified scenarios.
READ ALSO
The risk: confusing automation and control
Let's be clear. Connecting Finthesis to Pennylane is not enough to achieve quality financial management.
If the accounting plan is poorly structured, if analytics are absent, if flows are not properly categorized, if forecasting assumptions are weak, the tool will produce faster analyses, not necessarily more accurate ones.
This is where Blendy's support becomes crucial . Before discussing AI, it's essential to ensure that financial data is usable. This means clarifying revenue categories, tracking the right margin items, distinguishing between fixed and variable costs, ensuring the reliability of bank transactions, defining recurring expenses, structuring analytics, and selecting truly useful indicators.
A SaaS company isn't managed like an IT services company. An international e-commerce business isn't managed like a multi-location restaurant. A growing French SME isn't managed like a company already operating in France, Canada, and the United States.
Finthesis provides the platform. Blendy helps to build the financial analysis that corresponds to the business model.
How Blendy supports the use of Finthesis
Blendy 's support isn't about "installing a tool" and then leaving the company to fend for itself with its graphs. The value lies in the configuration, the methodology, and the interpretation.
Blendy can intervene from the structuring of the accounting foundation in Pennylane.
The goal is to feed clean, consistent, and sufficiently detailed data into Finthesis to produce useful analysis . This relies on the quality of data flows, the consistency of the chart of accounts, analytical management, reconciliation rules, and the reliability of historical information.
Next comes the choice of indicators. Too many charts stifle effective management. A growing SME needs a limited number of reliable signals: revenue, gross margin, payroll, cash flow, working capital requirement, profitability by activity, burn rate for certain models, average order value or recurring revenue as appropriate.
Blendy can also support the implementation of scenarios : hiring, investment, internationalization, launch of a new offer, opening of a foreign entity, price evolution, variation of payment terms.
This is often where Finthesis becomes truly interesting. The leader no longer just looks at the past. He tests different trajectories.
Finally, Blendy acts as a filter . The AI can produce a written analysis. It can highlight a key point. It can accelerate a business plan.
But it doesn't always know the commercial, human, fiscal, or international realities of the company. The consulting firm therefore retains a crucial role : verifying, interpreting, challenging, and transforming analyses into decisions.
READ ALSO
For which companies does Finthesis AI become truly relevant?
Finthesis AI becomes particularly interesting for companies that have already moved beyond the stage of simple accounting follow-up.
This is the case when a manager starts asking recurring management questions: Can I recruit now? Is it the right time to open a new market? Is my margin decreasing due to pricing, purchasing, staffing, or customer mix? How many months of actual cash flow do I have ahead of me? What scenario should I present to my bank or investors?
For an IT services company, the issue may concern the margin per mission, the payroll, the daily rates, the inter-contract period or recruitment.
For a SaaS, the financial analysis must include recurring revenue, growth, acquisition costs, profitability, and sometimes financing needs.
For an eCommerce business, the analysis must take into account payment flows, inventory, margins, currencies, and logistics costs.
For a growing restaurant business, the topic may concern material costs, payroll, profitability per point of sale and cash flow.
In all these cases, Finthesis provides a layer of analysis. Blendy provides the accounting and financial framework that makes this analysis reliable .
The proper use of financial AI
AI applied to finance is of little value if it merely writes pretty comments. It becomes useful when it helps to identify anomalies more quickly, formulate a diagnosis, prepare a scenario, and initiate a more precise discussion.
Finthesis is moving in this direction. The platform highlights intelligent analysis, annotated reports, natural language business plans, Claude integration via MCP, and industry benchmarks via Perplexity . It also emphasizes confidentiality, voluntary module activation, and user control. According to information published by the company, no confidential data is transmitted outside of Finthesis without explicit consent.
For Blendy , the benefit is clear: Finthesis can become an excellent component of the finance stack for SMEs looking to move from compliance accounting to performance accounting. Provided, of course, that priorities aren't reversed. The tool comes after the methodology. AI comes after clean data. Reporting comes after understanding the business model.
It is this combination that creates value: Pennylane to structure accounting, Finthesis to give depth to management, Blendy to transform figures into useful decisions.
Conclusion
Finthesis AI responds to a very concrete expectation of leaders: to understand more quickly what their figures tell us.
The promise is not to replace the accountant, the CFO, or the CEO. The real value lies elsewhere: making financial analysis more accessible, more responsive, and more decision-oriented.
For growing SMEs, particularly in digital services, IT services, SaaS, eCommerce, or restaurants, this type of tool marks a significant milestone. Financial data should no longer be confined to exports, spreadsheets, or reports read too late. It should inform day-to-day decision-making.
With Finthesis, Pennylane, and Blendy's support , the finance stack becomes more transparent. And above all, it begins to speak the language that truly matters to a leader: the language of decisions that need to be made now.
Sources:
With Blendy, an international digital CPA based in Paris, Montreal and Miami,, take advantage of all the benefits of digital accounting to accelerate your financial processes and grow your business.
Certified by Pennylane, Dext, QuickBooks and Stripe, we support digital, eCommerce, IT services, and SaaS companies in France and internationally.
