Skip to content
NILEROBOTAcademy
Applications close Sunday 1 November

Become a product engineer

Two small-group tracks: 12 weeks in Cairo for final-year students and fresh graduates, or 8 weeks online for developers already working full-stack. Ship real products with AI agents and get ready for teams abroad.

From Nilerobot, a software studio in Egypt. You learn the same spec, build, review and test workflow we use on our own ventures.

Apply for cohort 1Compare the two tracks

Why this cohort

The job is changing. Train for the new one.

More and more teams hire engineers who own a feature from the user's problem to production, and build it with AI agents. That's a Product Engineer. Its close cousin, the Forward Deployed Engineer, does the same work inside a customer's team.

Most postings for these roles ask for 2 to 5 years of experience, and some companies hire junior and graduate engineers. We teach the skills and help you run your own job search. We don't find you a job.

Compare

One cohort, two tracks

Same admission, same fee, and both end with the Product Engineer certificate. Choose the track that fits where you are now.

12 weeks · 12 seats

Foundations

Final-year students and fresh graduates

When
Four evenings a week: two in person in Cairo and two online. We agree the days and times with each cohort before classes start.
Your time
About 16–18 hours a week, 120 class hours
What you do
Rebuild the fundamentals without AI, build full-stack products, then engineering with AI agents and a team capstone.
See the 12 weeks
8 weeks · 6 seats

Professional

Developers already working full-stack, about 1–5 years

When
Two evenings a week online, on work days, plus a weekly 30-minute one-to-one review. We agree the days and times with each cohort before classes start.
Your time
About 10–12 hours a week, 28 class hours and 8 one-to-one reviews
What you do
A fast recap of the full life cycle on one reference app, then Claude Code and harness engineering in depth, agents across the whole pipeline, and a capstone for a client.
See the 8 weeks

Outcomes

By graduation, you'll be able to

  • Solve problems like an engineer

    Break a problem down, design clean objects, pick the right data structure and test your code.

  • Ship a full product

    Talk to users, then build the backend, database and frontend, ship it with a link you can share and run it in production.

  • Work with AI agents properly

    Build the agent's harness, write the spec, let it build in small steps, review every change, and catch it when it's wrong.

  • Be job-ready for a team abroad

    A CV, LinkedIn and GitHub that get read, a job search that's running, and mock interviews with written feedback behind you.

Roadmap

Foundations track · 12 weeks

Professional track · 8 weeks

The 12 weeks

The 8 weeks

Fundamentals first, then products, then AI agents. Every week ends with something you ship. One stack throughout: TypeScript, Node, React and Next.js, PostgreSQL, Git.

For developers already building web apps at work: a fast recap of the full life cycle on one reference app, then Claude Code and harness engineering in depth, agents across the whole pipeline, and a capstone for a client.

  1. Weeks 1–3

    Foundations

    Problem solving, object-oriented design, data structures, algorithms and clean, tested code, in weekly sprints. You write every line yourself.

    No AI tools
  2. Weeks 4–7

    Full-stack products

    User interviews, APIs and security basics, databases and system design, React and Next.js, a usability test, deployment and running it in production. You code with AI and explain every line.

    AI pair-programming
  3. Weeks 8–10

    AI agents

    Coding agents in depth, harness engineering and a starter kit, agents that test, a pull-request loop with a human approval, and one AI feature in your app.

    AI agents on
  4. Weeks 11–12

    Capstone

    Teams of 2 or 3 turn a vague client brief into a product through the agent pipeline, handle a change request and an incident, and take a support shift.

    AI agents on
  5. Weeks 1–2

    The full life cycle

    A fast recap on one reference app: requirements and agile, design, tests, security, CI/CD and a deploy. The baseline every agent must keep green.

    AI pair-programming
  6. Weeks 3–4

    Engineering with AI agents

    Claude Code in depth, the spec, plan, build and review flow, then harness engineering: hooks, lint rules, verification loops and parallel agents.

    AI agents on
  7. Weeks 5–6

    Agents across the pipeline

    MCP, skills and plugins, a monorepo, agents that test, a pull-request loop with a human approval, and an AI-feature lab.

    AI agents on
  8. Weeks 7–8

    Capstone

    Teams of 2 or 3 take a client brief through discovery and a written scope to a release they use, with an AI feature, an incident and a support shift.

    AI agents on
  9. Graduation

See the full curriculum, week by weekWhat you get at the end

Curriculum

The full curriculum, week by week

Tap any module to open it: each one shows what you learn week by week, what you practise and what you ship. The first one is open to show you.

Foundations track · 12 weeks

Professional track · 8 weeks

weeks
12
weeks
8
live sessions, 2.5 h each
48
live sessions, 1 h 45 min each
16
class hours
120
class hours
28
project work a week
6–8 h
project work a week
6–8 h
shipped projects
3
one-to-one reviews
8
students at most
12
developers at most
6
Module 01, Weeks 1–3Foundations
No AI tools

Rebuild the basics every interviewer will test: problem solving, object-oriented design, data structures, algorithms and clean, tested code. Taught live, and you write every line yourself, so the fundamentals are really yours before AI comes in. From the first week the cohort works like a real team, in weekly sprints.

  1. Week 1

    Tools, problem solving and how teams work

    Engineering
    Terminal, Git, GitHub and VS Code. TypeScript basics: types, functions, control flow. How to break a problem into steps before coding.
    Product
    What a product engineer does: users, problems and outcomes, not tickets. Agile from day one: the cohort runs weekly sprints with a board, user stories and short stand-ups.
    Work skills
    Introducing yourself in English. Writing a clear message on Slack.
    You ship by the end of the week
    10 solved problems in a GitHub repo, each with a short note on the approach
  2. Week 2

    Object-oriented design

    Engineering
    Classes, interfaces, encapsulation, composition over inheritance and the SOLID ideas in plain words. Modelling a small domain such as a library or a clinic booking system.
    Product
    Reading a feature request and turning it into a data model.
    Work skills
    Asking good questions when stuck, and the 30-minute rule before asking.
    You ship by the end of the week
    A domain model in TypeScript with classes and interfaces, plus a diagram
  3. Week 3

    Data structures, algorithms and testing

    Engineering
    Arrays, maps, sets, stacks, queues, recursion, simple algorithms and Big-O in practice. Unit tests, debugging, clean code and naming.
    Product
    Code review: how to give and take feedback on a pull request.
    Work skills
    The daily stand-up: yesterday, today, blockers.
    You ship by the end of the week
    A tested solution set and the first no-AI checkpoint, on the whiteboard
Module P1, Weeks 1–2The full life cycle
AI pair-programming

You already ship code. These two weeks are a fast recap of the whole life cycle, from a client's request to a release you can roll back, on one reference app (Next.js, Node and PostgreSQL) that stays your project for all eight weeks. Before week 1 you get a short pre-work pack: a self-check, a reading list (Git, TypeScript, SQL, HTTP and Docker) and a starter exercise on the reference app, so the recap can move fast.

  1. Week 1

    From requirement to design

    Session 1
    A full-stack recap on the reference app: architecture, API design and the data model, then UI design basics with a small design system and accessibility.
    Session 2
    Agile as teams really run it: from a client request to user stories with acceptance criteria, estimates, a sprint board and a definition of done. The life cycle as a map: requirements, design, build, test, release and operate.
    One-to-one review and career
    One-to-one: your goals, your current work and your pre-work self-check. From this week your pull request descriptions, design notes and status updates are also reviewed for clear written English.
    You ship by the end of the week
    A one-page spec, the user stories and a design for one new feature of the reference app
  2. Week 2

    Quality, security and the DevOps baseline

    Session 1
    The testing pyramid: unit, integration and end-to-end tests with Playwright. CI/CD with GitHub Actions, environments, Docker and secrets, then logs, error tracking and an uptime check. Releasing and rolling back.
    Session 2
    Security best practices: the OWASP Top 10 for web apps, auth, input validation, dependency and secret scanning, and a quick threat model of your new feature. Security checks in layers, from your editor to CI. Code review standards.
    One-to-one review and career
    One-to-one: the first no-AI checkpoint, where you explain and change your own code without the agent. Career: your CV for Product Engineer roles.
    You ship by the end of the week
    The reference app with tests, security checks, a green CI and a deploy: the baseline every agent must keep green
Module 02, Weeks 4–7Building real products
AI pair-programming

Full-stack web, the way product engineers work: talk to users, build the API, database and frontend, keep it secure, ship it and run it in production. From week 4 you code with AI, but you must be able to explain any line you push. Code reviews check that.

  1. Week 4

    Backend, APIs and security basics

    Engineering
    HTTP, REST and JSON, and a Node API with validation and error handling. Security basics: login, the OWASP Top 10 in plain words and keeping secrets out of your code. First steps with Claude Code: a learning mode that leaves parts of the code for you to write, asking it to explain a file and where it is called from, and model training turned off on your own account.
    Product
    Talking to users: a short interview with two real users, then a one-page spec from what you heard: problem, users, scope and what is out.
    Work skills
    Estimating work, and saying early when you will be late.
    You ship by the end of the week
    Notes from two user interviews, a one-page spec and a working REST API with tests and input validation
  2. Week 5

    Databases and system design

    Engineering
    SQL and PostgreSQL, data modelling, relations, migrations, indexes and an ORM. Your first system design: how data flows through your app, from the screen to the database and back.
    Product
    Choosing the smallest version worth shipping (the MVP).
    Work skills
    Writing a weekly async update for a manager abroad.
    You ship by the end of the week
    Your API backed by a real database, and a diagram of how data flows through it
  3. Week 6

    Frontend and product design

    Engineering
    React and Next.js: components, forms, state and calling your API. UI design basics: a small design system, and accessibility from the start.
    Product
    User flows and simple wireframes. A UI review: layout, hierarchy, empty, loading and error states, and use on a phone. Then a usability test with five people, and fixing what you saw.
    Work skills
    Demoing your work in 3 minutes.
    You ship by the end of the week
    A full-stack app a user can click through, tested with five people
  4. Week 7

    DevOps, system design and ship it

    Engineering
    Environment variables and secrets, Docker basics, environments and deployment, and GitHub Actions running your tests. Running it in production: logs, error tracking, an uptime check, and rolling back a bad release. How a simple system grows: caching, queues, a CDN and the trade-offs. Then a security review of your own app.
    Product
    Measuring whether a feature works: analytics events and one simple metric you check after release.
    Work skills
    Taking critical feedback. Career track starts: your CV, one page, projects first, numbers and links.
    You ship by the end of the week
    Your app deployed with a live link, logs, error tracking and analytics, its security review, the second no-AI checkpoint and CV version 1
Module P2, Weeks 3–4Engineering with AI agents
AI agents on

How coding agents really work, and the harness that keeps their work reliable when you use them a lot. You learn the patterns that carry across tools (instruction files, skills, MCP and hooks work in most coding agents), then go deep on one: Claude Code. Each week adds one habit that keeps the token bill low.

  1. Week 3

    Claude Code in depth, then the agentic flow

    Session 1
    How a coding agent works: a model, tools and context in a loop. Context engineering: why context is the scarce resource and how cost follows from it, with real prices. Choosing the model and effort, short instruction files that work as a map, permission modes and the sandbox, plan mode, and steering a long session. Token habit: pick the model and effort at the start, watch your usage and context, and clear between unrelated tasks.
    Session 2
    Spec-driven development with GitHub Spec Kit, as the basic agentic flow: once per project, a constitution (the rules every feature follows: stack, tests, security); then for each feature, specify (what and why; the agent interviews you), plan (how), tasks, implement, and converge until it reports done, with clarifications and checklists when a feature needs them. Fresh sessions between steps; tests verify; you review; then a pull request. The same idea works in other agents. A short debate on the usual objections ("it writes bad code", "it's slower"), and one tedious task timed both ways.
    One-to-one review and career
    One-to-one: your first agent sessions, reviewed with their usage numbers. Career: LinkedIn and GitHub for teams abroad.
    You ship by the end of the week
    The reference app's constitution, and week 1's feature built through Spec Kit (spec, plan, tasks and code) with a cost note
  2. Week 4

    Harness engineering in depth

    Session 1
    What happens when you use agents a lot, and the harness that keeps it under control: the instruction file as a table of contents and your docs, with the constitution and specs from week 3, as the system of record, hooks that enforce the rules in the constitution, lint rules whose messages tell the agent how to fix the problem, tests and CI as guardrails, and verification loops: a reviewer agent that checks the diff against the plan, and a verify step the agent runs before every commit. Long tasks with a feature list and a progress file. Token habit: noisy work goes to subagents on a small model, you compact only at natural breaks, and you know what resets the prompt cache.
    Session 2
    Parallel agents: subagents, git worktrees and cloud sessions. Git for agent work: rebasing, conflicts, bisect and reading big diffs. Reviewing AI-written code like a senior engineer, and removing harness parts newer models no longer need. A demo of Entire, which saves each agent session next to the commit it produced (private repos only, since the sessions are stored in the repo).
    One-to-one review and career
    One-to-one: your harness, reviewed. Career: the job search, target companies and short messages to hiring managers.
    You ship by the end of the week
    A harness for the reference app (hooks, a lint rule and a reviewer agent) and two features built in parallel in separate worktrees
Module 03, Weeks 8–10Engineering with AI agents
AI agents on

The skill teams abroad value most: running coding agents well at work. How agents work, the spec, plan, build and verify flow, harness engineering, agents that test and a pull-request loop where a human approves every merge, plus one AI feature in your app. You learn the patterns first, so they still hold when the tools change.

  1. Week 8

    Claude Code in depth and the agentic flow

    Engineering
    How a coding agent works: a model, tools and context in a loop, and why context is the scarce resource. Choosing the model, what it costs, and the first habits that keep the bill low. Then the flow, spec first with GitHub Spec Kit: a short constitution for your project (the rules every feature follows), then for one feature specify (what and why; the agent interviews you), plan, tasks and implement in small steps; tests verify and you review. Undoing a wrong turn, and catching the agent when it is wrong.
    Product
    Writing specs an agent can follow: from week 4's one-page spec to a Spec Kit spec.
    Work skills
    How product teams abroad work day to day. Career: your LinkedIn and GitHub, with a headline, featured projects and pinned repos with READMEs and live links.
    You ship by the end of the week
    A feature built by the agent through Spec Kit, with its spec, plan, tasks and review notes, and your LinkedIn and GitHub profiles live
  2. Week 9

    Harness engineering and the starter kit

    Engineering
    Harness engineering: the setup around a coding agent that makes its work reliable. Instruction files (CLAUDE.md, AGENTS.md), hooks, and tests, linters and CI as guardrails the agent must pass, so the rules in your constitution are checked, not just written down. Then a starter kit of MCP servers, skills and plugins that you learn to use well, checking a plugin before you install it, and keeping the agent safe: prompt injection, secrets and least privilege.
    Product
    Trust and responsibility: what the agent may do on its own and what always needs a human. You own every line it writes and can explain it, you say how you used AI, you check the licences of the code and packages it brings in, and decisions that affect people stay with people.
    Work skills
    Working across time zones. Career: the job search, where remote and foreign roles are posted, picking targets, a tracker and short messages to hiring managers.
    You ship by the end of the week
    A harness for your project with the starter kit set up, a list of 30 target companies and a tracker
  3. Week 10

    Agents that test, and the pull-request loop

    Engineering
    Agents in QA: a failing test first, then the fix, and Playwright test agents that plan, write and repair end-to-end tests. The pull-request loop: a review agent and CI comment, the agent fixes, and a human approves every merge. Git for agent work: rebasing and resolving conflicts.
    Product
    The AI-feature lab: one AI feature in your app, with any model API (bring your own Claude key if you want to use Claude). Where AI helps a user, where it gets in the way, and a few real test cases that show it works and treats different users fairly. The ethics of an AI feature: sending the model only the user data it needs, telling users when AI is involved, and trying prompt injection against your own feature.
    Work skills
    Writing a clear pull request description in English. Career: behavioural interviews with the STAR method, five stories from your projects, and questions to ask the interviewer.
    You ship by the end of the week
    Your pull-request loop with tests written by the test agents, an AI feature in your app, five written STAR stories and your first 10 applications sent
Module P3, Weeks 5–6Agents across the pipeline
AI agents on

Agents beyond your editor: the tools you give them and how to keep them safe, a monorepo they can work in, agents that test, and a pull-request loop that runs by itself while a human still approves every merge. You also build an AI feature into the reference app.

  1. Week 5

    Sharpening the agent: MCP, plugins, skills and monorepos

    Session 1
    MCP servers versus command-line tools plus skills, and when each costs fewer tokens. A starter kit you install and learn to use well. Writing your own skill, tested with it on and off, and a small MCP server. Checking a plugin before you install it, and agent security: prompt injection, the supply chain and least privilege (the OWASP Top 10 for LLM and agentic apps). Token habit: short instruction files, workflows moved into skills, and checking what each plugin adds to every turn.
    Session 2
    Monorepos for agents: one context, the project graph, affected tasks and generators, with Nx and Turborepo. You move the reference app into a monorepo with a shared library. Then Mohamed's own setup, shown live.
    One-to-one review and career
    One-to-one: your skill and its test. Career: system design interviews for AI products.
    You ship by the end of the week
    Your own skill with its on and off test, a small MCP server, and the app in a monorepo where the agent runs only the affected tasks
  2. Week 6

    Agentic PR review, the feedback loop and QA agents

    Session 1
    Agents in QA: a failing test first, then the fix; Playwright test agents that plan, write and heal end-to-end tests; performance and accessibility checks with Lighthouse; and evals for anything an AI produces, starting with 20 real cases and re-run in CI whenever the agent setup changes. Then the AI-feature lab: an AI feature in the reference app with structured output and its own evals, including cases that check it treats different users fairly; it sends the model only the data it needs, tells users when AI is involved and is tested against prompt injection. Use any model API (bring your own Claude key if you want to use Claude). With your own key you can also build a small agent of your own with the Claude Agent SDK.
    Session 2
    A pull-request loop that runs by itself: an agent opens the PR, a review agent and CI comment, the agent fixes, and a human approves the merge, with code owners and branch rules so an AI review never counts as the approval. Responsibility: you own every line an agent writes and can explain it, you say how AI was used, you check the licences of generated code and new packages, and decisions that affect people stay with people. A prompt-injection drill against your own reviewer. Agents in CI with turn and budget limits, agents on incidents through error tracking, and observability for AI output. Token habit: hooks that trim output to the failing tests, and budget caps in CI.
    One-to-one review and career
    One-to-one: the second no-AI checkpoint, and a look at your usage numbers. Career: behavioural interviews with STAR stories from your work.
    You ship by the end of the week
    The full loop running on your repo, end-to-end tests written by the test agents, and an AI feature with its evals
Module 04, Weeks 11–12Capstone and graduation
Team project

In teams of 2 or 3 you build one product for a simulated foreign client, with us playing the client: a vague brief you turn into a scope and a spec yourselves, two sprints through the agent pipeline from weeks 8 to 10, stand-ups in English, a change of requirements halfway, an incident to handle, a support shift and a release.

  1. Week 11

    Sprint 1

    Engineering
    Set up your team's shared harness and write its Spec Kit constitution and first specs from the agreed scope, then build the core flow through the pipeline: spec, agent, tests, a review agent and a human approval on every merge, with code review across teams. An incident drill: we break something in production, you find it from the logs, roll back or fix it, and write a short incident note.
    Product
    Starting from a vague brief: interview the client, decide what to build first and what to leave out, agree the scope in writing, and agree what the client lets you share with AI tools.
    Work skills
    Career: technical interviews. Live coding while thinking aloud, take-home tasks, explaining your own code, and when AI is allowed in an interview. Mock interview 1.
    You ship by the end of the week
    A working core flow, deployed, and your incident note
  2. Week 12

    Sprint 2 and graduation

    Engineering
    Handle the change request, polish, test and release. A support shift: answer the client's users and fix what they report.
    Product
    Release notes, the first numbers from your analytics, and a short case study.
    Work skills
    Career: mock interview 2, a full loop, recorded, with written feedback. Then offers, contracts and salary basics. Graduation day, in person, where you present your product.
    You ship by the end of the week
    A released product, a case study, your demo and your final career pack
Module P4, Weeks 7–8Capstone and graduation
Team project

In teams of two or three you build one product for a simulated client, from their brief to a release they use: discovery and a written scope, the whole agent pipeline from weeks 3 to 6, an AI feature, an incident and a support shift on the way, and two written documents the client keeps.

  1. Week 7

    Discovery, scope and sprint 1

    Session 1
    Build the core with the whole pipeline from weeks 3 to 6, connected to the client's systems and with an AI feature (any model API), and code review across teams. An incident drill in production: find it from the logs, roll back or fix it, and write the incident note.
    Session 2
    The client brief: a discovery session to map their workflow, then the scope agreed in writing, which becomes the team's Spec Kit constitution and first specs. Client confidentiality with AI tools: what the client's contract allows, never pasting their data or secrets into tools they haven't approved, and the basics of GDPR and Egypt's data protection law (Law 151 of 2020). Drafts of two documents for the client: an AI diligence statement (how AI was used and what data it saw) and an adoption plan for their own team.
    One-to-one review and career
    One-to-one code review. Career: a recorded mock interview with written feedback.
    You ship by the end of the week
    A written scope, a working core deployed, and an incident note
  2. Week 8

    Change request, release and graduation

    Session 1
    Handle a change request, finish and release, then a support shift answering the client's users.
    Session 2
    Release notes, a short case study, and the first usage numbers turned into the next spec. Graduation day, where you present your product.
    One-to-one review and career
    One-to-one: your plan after graduation. Career: offers, contracts and pay for roles abroad.
    You ship by the end of the week
    A released product, the AI diligence statement and adoption plan for the client, a case study and your demo
Weeks 7–12Career track

It starts in week 7, when you have real projects to show, and you leave with your job search already running. We teach you how to find and win a job yourself. We don't place you.

CV

  • One page in English, projects before education
  • Each project: what it does, your part, the stack and a live link
  • Reviewed twice: version 1 in week 7, final in week 12

LinkedIn and GitHub

  • A headline and about section written for recruiters abroad
  • Your capstone and best projects featured with demos
  • Pinned repos with clear READMEs and a steady commit history

Finding the job

  • Where remote and foreign roles appear: LinkedIn Jobs, career pages and startup job boards
  • 30 target companies, a tracker and a weekly routine
  • Short messages to hiring managers that get answers
  • Spotting fake jobs and unpaid "trial work"

Behavioural interview

  • The STAR method: situation, task, action, result
  • Five stories from your own projects
  • Talking about failure, conflict and learning

Technical interview

  • Live coding while thinking aloud
  • Take-home tasks and explaining your own code
  • Using AI in interviews only where it is allowed

Mock interviews and offers

  • Two mock interviews with written feedback, the second recorded
  • Reading an offer: employee or contractor, notice and payment
  • Salary basics: research the range, and never give the first number too early

Your career pack at graduation

  • Final CV
  • Live LinkedIn and GitHub profiles
  • A portfolio of 3 shipped projects with a demo video
  • Five STAR stories
  • A job search tracker with at least 10 applications sent
  • Two mock interviews with feedback
  • A certificate and a written reference
In your one-to-one, weeks 2–8Career track

Career work runs through your weekly one-to-one with Mohamed, next to the job you already have. We teach you how to find and win the next role yourself. We don't place you.

CV

  • Rewritten for Product Engineer roles
  • Week 2, in your one-to-one

LinkedIn and GitHub

  • Written for teams abroad
  • Week 3

Finding the job

  • Target companies and short messages to hiring managers
  • Week 4

System design interview

  • Designing AI products out loud
  • Week 5

Behavioural interview

  • STAR stories from your own work
  • Week 6

Mock interview and offers

  • A recorded mock interview with written feedback (week 7)
  • Offers, contracts and pay for roles abroad (week 8)

Your career pack at graduation

  • A CV for Product Engineer roles
  • LinkedIn and GitHub written for teams abroad
  • Your capstone product and its case study
  • STAR stories from your own work
  • A target list and messages to hiring managers
  • A recorded mock interview with feedback
  • A certificate and a written reference
Weekly, plus two checkpointsHow you're assessed

Weekly deliverable

Graded every week on four things: does it work, does it come with its proof (tests, screenshots or a reviewer's report), is the code clean, and can you explain it.

No-AI checkpoints

End of weeks 3 and 7: a problem on the whiteboard, no laptop. Using AI well depends on knowing what good code looks like without it.

Final grade

Deliverables 40%, checkpoints 20%, capstone 25%, work skills and attendance 15%. Graduates get a Nilerobot certificate of completion and a written reference.

Complete the work, no gradeHow you're assessed

Weekly deliverable

Reviewed in your one-to-one: does it work, does it come with its proof (tests, screenshots or a reviewer agent's report), is the code clean, and can you explain what the agent wrote. In weeks 2 and 6 a short no-AI checkpoint: explain and change your own code without the agent. No marks: you get feedback and fix what needs fixing.

Capstone

Built in a team for a simulated client, with a written scope, an AI diligence statement and an adoption plan, and presented on graduation day.

Certificate

No grade: complete the weekly work and the capstone and attend the sessions, and you get the Nilerobot certificate of completion for the professional track and a written reference.

One stack, start to finishTools you'll use

TypeScript everywhere, so you go deep instead of switching languages. It is also what most startups abroad that hire product engineers use.

  • TypeScriptone language, start to finish
  • Nodebackend and APIs
  • React and Next.jsfrontend
  • PostgreSQLdata
  • Git and GitHubevery day
  • VS Codeeditor
  • Claude Codecoding agent, from week 4, in depth from week 8
  • GitHub Spec Kitspec-driven development: a constitution, then specify, plan, tasks and implement, from week 8
  • Agent starter kitMCP servers, skills and plugins, from week 9
  • Playwrightend-to-end tests and test agents, week 10
  • Model APIsOptional: the AI-feature lab works with any model API, or your own Claude key with a spending limit you set
Your stack, plus the agent's kitTools you'll use

The stack you already work in, on one reference app. The new part is the tooling around it: a coding agent with a proper harness, a starter kit of MCP servers, skills and plugins, test agents and evals.

  • TypeScriptyour stack, start to finish
  • The reference appNext.js, Node and PostgreSQL, your project for all eight weeks
  • Git and GitHubevery day, with CI, worktrees and code owners
  • Claude Codecoding agent, in depth from week 3
  • GitHub Spec Kitspec-driven development, from week 3 to the capstone
  • Starter kitContext7 for current docs, GitHub read-only at first, Playwright, and plugins for commits, code review, type checks, frontend design and security, week 5
  • Only with carepayment, Firebase and cloud servers on a sandbox or dev project only, and memory plugins with a local provider
  • Nx and Turborepomonorepos for agents, week 5
  • Model APIsOptional: the AI-feature lab works with any model API, or your own Claude key with a spending limit you set

Questions about the curriculum? Ask us in your interview.

Certificate

Your certificate, signed

Every graduate who completes the cohort receives a certificate like this one, issued by Nilerobot.

More than three missed sessions without a reason means no certificate.

  • Signed by Mohamed Salah, CEO of Nilerobot and your instructor
  • Issued by Nilerobot, a registered Egyptian company, Commercial Registry No. 288635
  • Lists the programme, your track, its class hours (120 on foundations, 28 on professional) and what you studied
  • Has its own verification code: anyone can check it online, and you can add it to your LinkedIn profile in one click
  • Comes with a written reference you can share with employers
Sample
NILEROBOT Academy
Certificate no. NRA-C1-001

Certificate of completion

This certifies that

Your Name

has successfully completed

Product Engineer

12 weeks and 120 class hours of software engineering foundations, full-stack product development, engineering with AI agents and a team capstone, presented on graduation day.

8 weeks and 28 class hours of product engineering, harness engineering and AI agents in production, client delivery and a team capstone, with weekly one-to-one code reviews.

Mohamed SalahCEO, Nilerobot
February 2027January 2027Cohort 1Date of completion

Nilerobot, Commercial Registry No. 288635Mindhaus, District 5, New Cairo, Egyptnilerobot.comVerify at academy.nilerobot.com/verifyVerification code 7KQ2M-9XW4F

After graduation

What happens after you graduate

Graduation day is the finish line of the cohort and the start line of your career. Here is the path from there.

  1. Last week of your track

    Graduation day

    You present your capstone product and graduate.

  2. On graduation day

    You take home

    • A signed certificate of completion
    • A written reference
    • Your full career pack
    1. Some graduates, by invitation

      Internship at Nilerobot

      Keep building on real products with the team that taught you.

    2. Some graduates, by invitation

      Nilerobot's staff augmentation team

      Join as an engineer, working with companies abroad.

    3. Some graduates, by invitation

      An introduction to our recruitment partner

      They may put you forward for roles with companies abroad. It costs you nothing.

    Every graduate runs their own job search, with the career pack. Internships, team roles and introductions are never guaranteed, free for you and not part of what you pay for.

Schedule

A week in the cohort

Foundations track · 12 weeks

Professional track · 8 weeks

Four evenings a week, two in person in Cairo and two online, so final-year students don't miss university. We agree the days and times with each cohort before classes start, and online sessions are recorded. Plan for about 16–18 hours a week: 10 hours of live sessions and 6–8 hours on your project.

Two evenings a week online, on work days, recorded for days when work runs late. We agree the days and times with each cohort before classes start. Plan for about 10–12 hours a week: 3.5 hours of live sessions, a 30-minute one-to-one and 6–8 hours on your project.

In personBuild

A new concept, live coding and pair work.

Online, eveningBuild

The stack, the agent and its harness, live on the reference app.

OnlineBuild with AI

Build with AI agents on your own project.

Online, eveningProcess

Agile, security, the review loop and work with the client.

OnlineProduct

Specs, users and code review.

30 minutesOne-to-one

Your weekly code and career review with Mohamed, at a time we agree.

In personWork skills

Work skills and the career track.

Any timeRecordings

Every session is recorded, so a late day at work doesn't mean missing it.

At homeShip

Ship the week's work, with an hour of online questions.

6–8 hoursYour own time

Ship the week's work on your project, around your job.

Mohamed Salah

Lead instructor and mentor

Led by Mohamed Salah

CEO of Nilerobot, a software studio in Egypt and a Google Cloud Partner Advantage member. He leads engineering on production software for US companies and builds with AI agents every day. He teaches the cohort, reviews your code and mentors you, and the academy teaches the way his team works.

Mohamed Salah on LinkedIn (opens in a new tab)Nilerobot on Google Cloud (opens in a new tab)

Mohamed Asmaay

Testimonial

I started working at Nilerobot, and a few months later I began working with a client in Europe. It was new for me, and a great experience. I learned a lot working alongside Mohamed, and his mentoring is what helped me do well with clients abroad.
Mohamed AsmaayProduct EngineerMohamed Asmaay on LinkedIn (opens in a new tab)

Admission

How to join

Twelve seats on the foundations track and six on the professional track, chosen for good English, solid fundamentals, clear reasoning and the right attitude.

  1. 1 NovApplications close
  2. 10 NovTests and interviews done
  3. 12 NovDown payments due
  4. 15 NovFirst class (professional: 16 Nov)

  1. Apply, free

    A 5-minute form: your track, your studies or work, your national ID and your CV or GitHub.

  2. Online step, free

    About 45 minutes in your own time: a 30-minute quiz, a 60–90 second English video and a short written answer.

  3. Shortlist: take-home

    A small project, AI tools allowed: up to 4 hours over 3 days (professional: 6 hours over 4 days).

  4. Interview

    30 minutes: walk us through your take-home and make one small change live. Everyone interviewed gets written feedback.

  5. Offer and down payment

    Pay the 3,500 EGP refundable down payment to hold your seat. The rest is monthly, from your first class day.

Included

  • Every live session, recorded
  • Weekly code review
  • Capstone and graduation day
  • Career coaching and mock interviews
  • Certificate and a written reference

You bring

  • Your own laptop
  • A Claude Pro subscription on your own account, about $20 a month
15,500 EGPfounding price, both tracks
  1. Down payment, after acceptance, refundable until the startBy 12 Nov3,500
  2. November installmentDue 15 Nov4,000
  3. December installmentDue 1 Dec4,000
  4. January installmentDue 1 Jan4,000
  5. Total15,500 EGP
  1. Down payment, after acceptance, refundable until the startBy 12 Nov3,500
  2. November installmentDue 16 Nov6,000
  3. December installmentDue 1 Dec6,000
  4. Total15,500 EGP

Full scholarship, foundations track

The top applicant studies free

No down payment and no monthly payments for the best combined test and interview score, decided on the scores alone.

This is a training programme. It doesn't include job placement.

FAQ

Questions

What happens after graduation?

Nilerobot runs a staff augmentation team: its own engineers working with companies abroad. It sometimes takes graduates on as interns or as engineers in that team, or introduces them to its recruitment partner, at no cost to them. It's never guaranteed and it isn't part of what you pay for. Everyone leaves with the skills and career pack to find a job themselves.

Which track is for me?

Foundations if you're in your final year or graduated recently and want strong fundamentals first. Professional if you already work as a full-stack developer, about 1 to 5 years, and want the AI, product and client skills. You choose in the form, and we can suggest the other track after the interview.

I work full-time. Can I join?

Yes, on the professional track: two evenings a week online on work days (recorded for days when work runs late; we agree the days and times with each cohort), a 30-minute one-to-one review, and 6 to 8 hours on your project. It is built around a job. The foundations track meets four evenings a week, two of them in person, which is very hard next to a full-time job.

Why does the professional track cost the same?

It is shorter, but the class is six developers, it covers the basics in a fast two-week recap and spends the rest on the senior material (Claude Code and harness engineering in depth, agents across the pipeline, client work), and every week you get a one-to-one code and career review with Mohamed.

Do I need to know how to program already?

Yes, the basics. You should be able to write simple programs in any language. On the foundations track the first three weeks rebuild the fundamentals properly, but they don't start from zero. The professional track skips them: it is for developers already working full-stack.

Where are the in-person sessions?

On the foundations track, in Cairo. The professional track's classes are online. We share the exact venue with accepted students before the cohort starts.

What language is it taught in?

English, with Arabic when it helps understanding. Working in English is part of what you practise.

Is applying free?

Yes, and so is every step before an offer: the form, the quiz, the English video, the take-home and the interview, with written feedback if you're interviewed. You pay only if you're offered a seat and take it: a 3,500 EGP refundable down payment, then the monthly installments.

Can we pay monthly?

Yes. After a 3,500 EGP refundable down payment, the rest is paid monthly from your track's first class day. Foundations: three installments of 4,000 EGP (15 November, 1 December and 1 January). Professional, which is shorter: two of 6,000 EGP (16 November and 1 December), so it's all paid before the track ends. Either way the total is 15,500 EGP, and applying costs nothing.

How much time does it take each week?

Foundations: about 16–18 hours, with four live evening sessions of 2.5 hours and 6–8 hours on your project, built around daytime university. Professional: about 10–12 hours, with two evening sessions of 1 hour 45 minutes online on work days, a 30-minute review and 6–8 hours on your project, built around a job.

Is there a scholarship?

Yes, one full scholarship for cohort 1, on the foundations track. The applicant with the best combined test and interview score studies free: no down payment and no monthly payments. Everyone takes the same assessment, it's decided on the scores alone, and we announce it with the offers.

Why isn't the AI subscription included?

Each student needs their own personal Claude account, so their work and history stay theirs after the course. The Claude API is optional and billed separately, by use. The AI labs work with any model API, and if you'd like to use Claude's, you add your own credit and set a spending limit; we show you how to keep the bill low.

What if I miss a session?

Online sessions are recorded. For in-person sessions, you catch up with the week's material and your project. More than three missed sessions without a reason means no certificate.

Do you guarantee a job?

No. This is a training programme. You leave with real skills, shipped projects, a career pack, a certificate and a written reference. We teach you how to find and win a job yourself.

Apply for cohort 1

Choose your track and fill in the form. Applying is free, and you go straight on to the free online step: the quiz and a short English video. Applications close Sunday 1 November, or earlier if the seats are taken.

Start your application