The problem behind the product Anyone who works on data projects has probably been through this at some point: the conversation with the client is going well, the scope is starting to take shape, and then the most uncomfortable part of the meeting arrives. Price. "So now, how much should I charge? Will the client be scared by this number?" That is not random insecurity. Pricing a data project involves variables that almost never fit into a simple equation: integration complexity, number of sources, data maturity, governance requirements, predictive modeling, ML deployment in production... every layer changes effort, timeline, and perceived value. That pain and complexity are exactly why I decided to build Valora . Who Valora was built for The focus is on data professionals working as freelancers, consultants, or small consulting firms. In practice, the product is designed for three profiles: Data Analysts: people working with dashboards, visualizations, reports, and analysis, usually with BI tools or similar platforms. Data Engineers: people who design and build pipelines, ETL/ELT processes, data ingestion, and governance structures. Data Scientists: people working with statistical modeling, machine learning, forecasting, and GenAI solutions. Those profiles may look close to each other from the outside, but the scope changes a lot between them. Because of that, the pricing logic also has to change. It would make no sense to throw a generic formula on top and pretend the problem was solved. What Valora does In practical terms, the platform organizes four stages: 1. Smart pricing The calculator does not just ask, "How many hours will you work?" It considers the professional area, technical complexity, number of sources and APIs, maintenance requirements, data volume, and even risk factors. On top of that, the professional can adjust hourly rates, complexity weights, and effort distribution across phases. The result stops being a guess and becomes a realistic and fair estimate, including a projected timeline. Your browser does not support HTML5 video. 2. Commercial proposals in PDF Once the calculation is done, the proposal is generated as a PDF in seconds. The document already includes scope, deliverables, timeline, and pricing, all with customizable visual identity so it does not feel like an improvised attachment. On paid plans, the PDF is clean. On the free plan, the watermark stays, but without making testing impossible. The idea is to let users validate the flow before deciding to subscribe. Your browser does not support HTML5 video. 3. Contracts with an AI assistant This is one part I find especially interesting: instead of dropping a generic contract template into the workflow, the platform offers an AI assistant focused on legal and technology contracts. That makes it possible to generate clauses from the project description, review existing text, translate contracts for international clients, and maintain different contract profiles. The contracts remain editable, with the ability to add, remove, and reorganize clauses via drag and drop. The visual identity follows the proposal, which helps keep consistency across the documents. Your browser does not support HTML5 video. 4. Legally valid digital signature After generating the contract, you just send the link to the client's email so they can sign it. No heavy attachments, no scanning, and no need to install anything. The electronic signatures use SHA-256 hashing and follow Brazilian Law 14.063/2020 and Provisional Measure 2.200-2/2001. On top of that, the system records an audit trail with IP, date and time, verified email, and the signature image, preserving document integrity, which is crucial for compliance audits or legal disputes. Your browser does not support HTML5 video. The dashboard shows status in real time: sent, pending, signed, or expired. It also centralizes useful actions such as resending the link, downloading the final contract, and canceling requests. Supporting features Beyond the main flow, there is a set of features that supports day-to-day usage: Document history: proposals, contracts, and calculations are saved with filters by type, date, or ID, making it easier to review or reuse older material. Help meter: a dashboard showing generated volume, time saved, and a light layer of gamification to track consistency of use. Real-time notifications: alerts for views, signatures, platform updates, and payment events. AI credits: the assistant works through monthly credits on paid plans, covering clause generation, legal review, and translation. Adaptive interface: support for light mode, dark mode, and responsive navigation across desktop, tablet, and mobile. Pricing model The good part is that the pricing structure does not try to look more complex than it needs to be. The reading is almost immediate: one plan to get started, one to operate at a steady pace, and one to use without a low ceiling. Free R$ 0 2 pricing runs per month 2 proposals per month 1 contract per month Watermark on PDFs 3 AI credits as a trial Professional R$ 34,90 /month 30 pricing runs per month 30 proposals per month 20 contracts per month 2 digital signatures per month 15 AI credits per month Full history Calculator parameterization Visual customization Expert R$ 49,90 /month Unlimited pricing runs Unlimited proposals Unlimited contracts 20 digital signatures per month 65 AI credits per month Priority support Early access to new features Payments are processed via Stripe, with a secure portal to manage subscription, card, and invoices, while also staying aligned with consumer protection law for cancellations . Security and compliance The foundation includes secure authentication, protection against automated access, password recovery with temporary links, encryption, and practices aligned with Brazil's LGPD. There is also the option to request full deletion of the account and data. For digital signatures, the links expire automatically and each step generates an audit record with IP and date/time. This is the kind of detail that adds stronger legal confidence to contract execution. Under the hood: the calculator If I had to point to the most sensitive part of the codebase, it would not be the prettiest screen or the most marketable feature. It would be the pricing-calculator . That is where the application stops being just an interface and starts carrying business logic that actually has to make sense. Critical excerpt This is not a function that simply does hours x hourly rate and calls it a day. The pricing-calculator works like a small system with three specialized calculators, each with its own weights, rules, and exceptions. 3 specialized calculators 10 steps in the analyst flow 700 lines of configuration 23 Edge Functions calculatePricing() // routes by profile if area === 'analyst' -> calculateAnalystPricing() if area === 'engineer' -> calculateEngineerPricing() if area === 'scientist' -> calculateScientistPricing() The analyst calculator, for example, is not a single equation. It goes through a decision pipeline: initial analysis, infrastructure, modeling with volume multipliers, API integrations weighted by documentation quality, KPI definition, visual complexity, ETL optimization for large volumes, and a recurring support layer when maintenance exists. 01 Initial analysis 02 Infrastructure Included when the project requires additional technical groundwork. 03 Data modeling Considers volume and structural effort. 04 API integrations The weight changes according to documentation quality. 05+ KPIs, visuals, and support The calculation keeps adding nuance, not just hours. For engineers, variables such as pipeline type, number and type of sources, governance level, source data quality, and deployment format come into play. For scientists, the weight changes according to data maturity, project goal, delivery model, testing scope, and validation level. The challenge was not just to calculate. It was to calculate without over-simplifying work that is inherently full of nuance. Another point I really like is that these weights are not rigid. On paid plans, users can parameterize the logic. The calculator settings file has nearly 700 lines just for weights, labels, and descriptions. It is almost like adjusting an airplane dashboard: every tweak changes the output in real time. In other words The most complex module is not the one that shows up the most. It is the one that holds the product together. If this calculation were shallow, the entire value proposition would collapse with it. There is also a second code path worth mentioning: the digital signature. When the client signs, the frontend calculates the SHA-256 hash via crypto.subtle.digest , combining envelope ID, email, IP, and timestamp. The result becomes a unique 64-character hexadecimal hash stored with the document before the call to the finalize-signature Edge Function. ai-assistant clauses, legal review, and translation finalize-signature persists the PDF and closes the flow expire-old-envelopes automatic document expiration schedule-plan-change future plan changes + 19 functions credits, limits, analytics, and product routines On the backend, the application is spread across 23 Edge Functions. In short: it took serious work to structure, prototype, and test all of this without neglecting security. Why this matters in my portfolio This project represents a way of thinking about products: identify a real operational friction point, design the end-to-end experience, structure the logic behind the solution, and put it into production with payments, AI, and digital signature actually working. This is not an academic exercise or a conceptual landing page. It is a working SaaS product, with real users, active billing, legal signing, and a pricing engine solving a concrete problem. See the product If you want to see all of that in action, the link is right below. → See Valora →