Fora Soft

It starts with a discovery call where we clarify the business goal, users, and constraints. Within days you get structured requirements (SRS / PRD), our open questions, and flagged assumptions, so gaps surface before development, not during it. Then an analyst turns requirements into specs and a clickable prototype (Visily or custom HTML) you can edit yourself. Nothing goes to engineers until you confirm the specs match your vision.

We qualify the business goal before any estimate: what the product should earn or save, for whom, and by when. Early estimation includes competitor analysis for new products, so scope decisions are grounded in the market, not just the backlog. Specs are agreed with you by our analyst, and the clickable prototype lets you correct direction while changes are cheap. During development, monthly demos (or your preferred cadence) check the build against those goals rather than against a task list.

Time & Materials with an Agile, iterative flow is our default: it fits evolving requirements, legacy codebases, and R&D. Fixed Price is available when scope, timeline, and risks are fully defined up front, and runs on milestones with weekly written status reports. Since 2025, both models run on spec-driven agentic engineering: senior engineers orchestrate AI agents against agreed specs, review every line by hand, and code the complex parts themselves. That is what lets the same team ship 4-10x faster.

You always see the current state of the product: clients get continuous access to the latest version of the system, not a monthly build. You receive detailed weekly timesheets showing hours per task & written status reports.

Demos run monthly by default, or on whatever cadence you prefer: some clients want them weekly, some per milestone. Between demos you get weekly timesheets & status reports, day-to-day contact with the project manager, and continuous access to the latest version, so a check-in is never the only window into progress.

Every project has a test engineer, and delivery includes a dedicated stabilization and final testing phase before we call the system done. In agentic development, AI-assisted review runs on every pull request, and senior engineers still review every line by hand. We assess code against a 21-parameter checklist covering architecture, security, and maintainability, the same one we use to audit external codebases.

Risk starts with the engagement model: our decision tree defaults uncertain scope, legacy code, and R&D to Time & Materials, so estimates adjust instead of breaking. Estimates come per block in two values, most likely and pessimistic, and open questions and assumptions are flagged to you before work starts. Legacy systems get a technical analysis before any commitment.

Most clients keep us for maintenance: part-time ongoing work on improvements after launch. We deploy to your servers, help with app store submission, and hand over updated source code.