Question 1

How do I start a project with Ascii-Core?

Accepted Answer

The first step is a 30-minute discovery call. Book it at asciicore.com/book or send an email to rihab_nikh@outlook.fr. On the call, we discuss your business context, the problem you are solving and what a successful outcome looks like. We do not require a brief or specification before the call — most clients come with a rough idea and we help shape it into a concrete scope.

Question 2

What happens on the discovery call?

Accepted Answer

The discovery call is a two-way conversation. We ask about your current process, the pain it creates and what success would look like. We share relevant experience from similar projects and give an initial sense of feasibility, approach and rough timeline. There is no sales pressure — if we are not the right fit, we will say so and recommend someone who is.

Question 3

How does pricing work?

Accepted Answer

We price projects as fixed-fee engagements based on the scope agreed during the Discovery phase. There are no hourly rates, no per-task charges and no change orders for in-scope work. If scope needs to expand, we discuss and agree the cost before any additional work begins. After launch, clients can choose a monthly retainer for ongoing support and feature development.

Question 4

What is a typical project timeline?

Accepted Answer

A focused AI agent or automation workflow typically ships in 4–6 weeks from the start of Discovery. A full web or mobile application is typically 8–12 weeks. Larger platforms combining AI, automation and custom software are delivered in phased milestones — most clients see working results within the first two weeks of development regardless of project size.

Question 5

What information do you need to get started?

Accepted Answer

Not much for the initial call. A rough description of the problem, the type of system you think you need and any relevant context about your existing tools and data. We conduct a formal technical audit during the Discovery phase. The more context you can share about your current process and pain points, the more productive the first conversation will be.

Question 6

What types of business processes can AI automation handle?

Accepted Answer

AI automation works best for repetitive, rule-based or pattern-recognition-heavy processes: document extraction and classification, lead qualification, customer support routing, data entry from emails or forms, report generation, appointment scheduling and CRM updates. Processes involving open-ended judgment calls still typically require human oversight, but AI can handle the structured portions and flag exceptions for human review.

Question 7

What data do I need to have in place before starting an AI project?

Accepted Answer

Less than most people think. For AI agents that interact with your existing systems (CRM, ERP, database), we use your live data via API — no migration required. For RAG systems that answer questions over your documents, we need your document library in any common format (PDF, Word, web pages). For custom-trained models, we discuss data requirements during Discovery and are honest if the available data is insufficient for the use case.

Question 8

How accurate are AI systems in practice?

Accepted Answer

Accuracy depends heavily on the use case and the quality of inputs. Document classification systems we have built achieve 90–97% accuracy on well-structured documents. Customer support routing runs at 85–95%. For mission-critical workflows, we always design human-in-the-loop review for low-confidence outputs — AI handles the high-confidence cases autonomously and routes edge cases to a human queue.

Question 9

What ongoing maintenance do AI systems require?

Accepted Answer

LLM-based systems require prompt updates when the underlying model is updated or when output quality drifts. Automation workflows require updates when the integrated systems change their APIs or data formats. We include one round of post-launch calibration in every project and offer retainer-based maintenance for ongoing monitoring and updates.

Question 10

Can n8n be self-hosted on our own servers?

Accepted Answer

Yes — that is our default recommendation. Self-hosted n8n runs on a VPS (typically $20–100/month) or within your existing cloud infrastructure. Data never leaves your environment. We handle the deployment, configuration and initial setup as part of the project. Ongoing hosting management is available as part of a retainer.

Question 11

Which LLM do you use — OpenAI or Anthropic?

Accepted Answer

Both, depending on the use case. GPT-4o is our primary model for agentic systems and function calling. Claude is preferred for long-context document analysis and situations where constitutional alignment reduces the need for extensive guardrails. For data-sensitive projects where no data should leave the organisation, we evaluate locally-hosted open-source models (Llama, Mistral) running on client infrastructure.

Question 12

Why do you use Next.js instead of WordPress?

Accepted Answer

WordPress is appropriate for primarily content-driven sites where the owner manages content without engineering involvement. Next.js is the right choice when the site integrates with APIs, requires AI features, has complex user interactions, needs high performance or will grow into a web application. For Ascii-Core clients building AI platforms, SaaS products or data-driven marketing sites, Next.js is consistently the better long-term investment.

Question 13

Can you integrate a CMS for non-technical content editing?

Accepted Answer

Yes. We integrate headless CMS platforms — most commonly Sanity, Contentful or Payload — with Next.js so marketing teams can manage content without engineering involvement. The CMS manages the content; Next.js handles rendering, performance and SEO. We recommend Payload CMS for most projects because it is open-source and self-hostable.

Question 14

How do you handle SEO for Next.js applications?

Accepted Answer

Next.js server-renders HTML, which search engines index directly without executing JavaScript. We implement the Metadata API for title, description, Open Graph and structured data (JSON-LD). Sitemaps are generated programmatically from the content database. Core Web Vitals are addressed through React Server Components, the Image component and route-level code splitting. Most Next.js sites we build score 90+ on PageSpeed Insights.

Question 15

Where do you host web applications?

Accepted Answer

Most Next.js applications are deployed on Vercel — optimal performance, preview deployments per pull request and zero-configuration Next.js optimisation. For clients with data residency requirements, budget constraints or preference for self-hosting, we deploy on Docker to Railway, Render, Fly.io or a dedicated VPS. The architecture supports migration between hosting providers.

Question 16

Do you provide ongoing maintenance after launch?

Accepted Answer

Yes. Post-launch maintenance is available as a monthly retainer covering dependency updates, bug fixes, security patches, performance monitoring and a defined number of hours for feature additions. Most clients choose a retainer — web applications require regular dependency updates, and having the team that built the system handling maintenance avoids knowledge transfer overhead.

Question 17

When do you use React Native and when do you use Flutter?

Accepted Answer

React Native is our default for most projects: one TypeScript codebase for iOS and Android, shared logic with the web frontend, and the Expo ecosystem for builds and OTA updates. Flutter is the better choice when pixel-perfect custom UI is required, when cross-platform coverage extending to web and desktop is needed, or when the team already uses Dart. We discuss framework choice during Discovery and are honest about the trade-offs.

Question 18

How much does a mobile app cost to build?

Accepted Answer

A single-feature MVP for iOS and Android typically starts in the low to mid five figures. A full-featured application with authentication, real-time data, payments and AI integration is scoped individually after Discovery. The cross-platform approach saves 40–60% compared to building separate iOS and Android apps with different teams.

Question 19

How long does App Store and Google Play approval take?

Accepted Answer

Apple App Store review takes 1–3 business days for new submissions and major updates. Google Play takes 1–3 days for new apps and under 24 hours for most updates. With Expo Over-The-Air updates, JavaScript-layer changes — bug fixes, UI tweaks, content changes — can ship to devices without a store review cycle.

Question 20

Can you update the app after launch without App Store submissions?

Accepted Answer

For JavaScript-layer changes, yes. Expo OTA updates allow us to push bug fixes, UI changes and content updates directly to installed apps within minutes. Changes that modify native code — new device API access, third-party native module additions — still require an App Store submission.

Question 21

Do you support Arabic and right-to-left layout in mobile apps?

Accepted Answer

Yes. We have built Arabic-first mobile applications for Moroccan and GCC clients. React Native has built-in RTL layout support. Arabic font rendering, bi-directional text and RTL-aware component layout are handled as part of the standard development process for projects with Arabic requirements — not as an afterthought.

Question 22

Do you run weekly demos during the project?

Accepted Answer

Yes — guaranteed. Every Friday we hold a 30–45 minute video call where we walk through what was built that week on a live staging environment. You interact with real software, not slides. A recording and written summary are shared after each session. Feedback from the demo is incorporated into the following week's work.

Question 23

Who owns the intellectual property and code?

Accepted Answer

You do. All code, designs, documentation and intellectual property created for your project are transferred to you on final payment. We retain no rights to your codebase, designs or business logic. This is written into every contract. You receive the full source code in a repository you control.

Question 24

What are the payment terms?

Accepted Answer

Typically 30% on contract signature, 40% at the midpoint milestone (working software on staging) and 30% on production deployment. For retainer engagements, payment is monthly in advance. We accept bank transfer and card payment. For Moroccan clients, we invoice in MAD; for international clients, in EUR or GBP.

Question 25

What happens if the project goes over scope or schedule?

Accepted Answer

In-scope work is delivered for the agreed price and timeline, regardless of how many hours it takes us. If scope needs to expand — new features, integration requirements discovered during development, significant specification changes — we flag it immediately, scope the addition jointly and agree a price and timeline in writing before any additional work begins. You never receive a surprise invoice.

Question 26

How is the project handed over at the end?

Accepted Answer

Handover includes: source code in a repository you own, a runbook covering deployment, configuration and common operational tasks, API documentation, architecture diagrams, a live walkthrough session with your team and 30 days of post-launch support included in the project fee. We can also provide an extended retainer for ongoing maintenance and development.

Question 27

Can you integrate CMI (Centre Monétique Interbancaire) for Moroccan payments?

Accepted Answer

Yes. We have integrated CMI for Moroccan card payments (Visa/Mastercard issued in Morocco) alongside international payment processors. CMI requires a merchant agreement with a Moroccan bank; we handle the technical integration once the merchant agreement is in place. For international clients also serving Morocco, we typically implement both CMI (for MAD transactions) and Stripe (for EUR/GBP/USD).

Question 28

How do you handle CNDP (data protection) compliance in Morocco?

Accepted Answer

Morocco's Law 09-08 on personal data protection, enforced by CNDP, requires data controllers to notify CNDP of personal data processing activities in certain categories. We document the data processing architecture, identify which processing activities require CNDP notification, and produce the technical documentation required for the notification process. Legal counsel on the regulatory interpretation is the client's responsibility.

Question 29

In which languages do you work?

Accepted Answer

We work natively in Arabic, French and English. Project communication, documentation and deliverables can be produced in any of the three languages or a combination. For clients where the end user base is Arabic-speaking, we design and develop with Arabic language support and RTL layout requirements from the start.

Question 30

Can we meet in person in Casablanca?

Accepted Answer

Yes. We are based in Casablanca and are available for in-person meetings for kickoffs, discovery workshops, project reviews and demos. For clients outside Morocco, we operate fully remotely with video calls, recorded demos and asynchronous communication that works across time zones.

Everything You Need to Know

Getting Started questions

AI & Automation questions

Web Development questions

Mobile Development questions

Delivery & Process questions

Morocco-Specific questions

Ask us directly

Ready to start a project?