Coodesh follows a transparent, sustainable pricing policy, designed to keep the service running with high availability, security and support, without straining the budgets of companies in Brazil and across Latin America. This article shows the real costs behind running a technical and behavioral assessment platform at scale, so you understand why our plans and prices are set the way they are.
The infrastructure behind every test
We keep data on MongoDB Atlas, with replicas, backups and auto-scaling to handle traffic spikes during hiring cycles. We run on AWS (compute, networking, observability) and use Amazon S3 to securely store files, assessment evidence, videos and reports, delivered through a content delivery network (CDN).
We also run our own code execution servers, which test candidates’ solutions against test cases in dozens of languages, plus environments for SQL and spreadsheet questions. Invitations and notices go out by email, SMS and WhatsApp. Each of these services charges for capacity, requests, data transfer, storage or messages sent, and the cost moves with usage and exchange rates.
AI at every step
AI powers a large part of what the platform delivers:
AI grading: open answers, code, essays, projects, whiteboard and audio and video answers are graded against criteria, with a score, feedback and an explanation of how each criterion was assessed.
AI interview: an interviewer holds the conversation by voice or text, in real time, and asks follow-up questions. Every turn of the conversation involves speech transcription, a language model and voice synthesis.
Audio and video transcription: recorded answers are transcribed with speaker separation, and the transcript is graded by AI.
Content creation and organization: generating questions, grading criteria, skills and job descriptions, translating the question library and reading résumés automatically.
We use state-of-the-art AI models from leading providers, with orchestration layers that uphold quality and privacy. AI isn’t free: there is a cost per token and per minute of audio processed, and different models for different tasks. In pricing, we factor in the average processing volume per customer and usage ceilings that keep things predictable.
Integrity and candidate identity
To make assessment results trustworthy, we offer features that confirm who is taking the test and how it was taken:
Webcam monitoring: periodic photos analyzed by facial recognition AI, which flags a missing face, more than one person in the frame, or looking away.
Screen recording: a record of what happens in the assessment tab, for later review.
Document verification and liveness check: confirm that the person taking the assessment is the person invited, by matching an official ID document against the face captured in real time.
Browser controls: tab switching, copy and paste, leaving full screen and use of a second monitor.
Background check: a report on the candidate built from external databases.
Several of these features rely on specialized providers that charge per check, on top of processing and storing images and recordings. That is why some of them have plan limits or use their own credits.
People and support
Beyond infrastructure, we provide real-time support tools in the product and a human support team dedicated to helping customers and candidates. Our team speaks Portuguese, Spanish and English, which reduces friction across Latin America and in global operations. We also have developers continuously improving the platform (new tests, ATS integrations, performance, security), which means salaries, tooling, QA and a reliable deployment pipeline.
Costs in dollars, prices in local currency
We are a Brazilian startup with no venture funding, competing with global players that often price in USD. Most of our inputs (cloud, AI, identity verification, monitoring) are also dollar-denominated. Our commitment is to absorb as much volatility as possible and keep pricing fair and stable, prioritizing long-term relationships.
Currency: in Brazil we invoice in Brazilian reais (R$). For companies outside Brazil, we charge in USD or adapt to local models in markets such as Mexico, Argentina, Colombia and Chile (for example, billing terms and methods that match local practice), always seeking a balance between predictability and accessibility.
Reliability has a cost
Running a product like ours is not simple. It takes high uptime, fast incident response, data protection, fraud prevention, elasticity for candidate peaks, and human support when the unexpected happens. Our pricing reflects the reliability standard customers expect during critical hiring and internal assessment moments.
Plans and price reviews
Our plans balance different needs: occasional usage, growing teams, and enterprise operations with steady demand and specific requirements. Each tier includes usage limits, features and support levels that optimize cost-benefit for your stage and predictability. Where possible, we offer monthly and annual options, with annual plans designed for budget predictability and a lower effective cost.
We periodically review pricing based on real usage, exchange rates, cloud and AI costs, and our roadmap. When we gain efficiency (infrastructure, models, caching, optimizations), we reinvest it in product improvements and price stability. When external costs rise, we consider adjustments with maximum caution, prioritizing service continuity and long-term partnerships.
Our commitment
We aim to deliver measurable value: shorter screening cycles, more accurate hiring, a better candidate experience and clearer skills visibility for fairer decisions. If something doesn’t fit your reality, we want to hear it. We’ll tailor the plan recommendation, guide your setup and help you get the most from your investment.
Gabriel Ferreira, CEO and Co-Founder
For further questions, contact gabriel@coodesh.com.