Discover
Understand the users, workflow, data and business baseline.
Build it. Deploy it. Explain its value.
Become a Forward Deployed Engineer
Turn a client-style business problem into a working AI system you can deploy, measure and hand over. Build across Python, APIs, interfaces, retrieval and agents—then show the engineering decisions behind your work.
The stack you build with
PythonFastAPIReact / TypeScriptOpenAI / ClaudeLangChain / LangGraphMCPPostgreSQL / pgvectorDockerGitHub ActionsAWS / Azure
12weeks · live instruction
18learning units
45hands-on labs
17checkpoint deliverables
1capstone & Demo Day
18 learning units = 12 technical modules + 6 Neriya client-delivery workshops. Cohort dates and live-session times are confirmed before enrollment.
Engineer · Architect · Consultant
Own the journey from problem to handover.
Practice the whole engagement, including the work around the code.
Design
Agree scope, acceptance criteria and integration architecture.
Build
Connect a full-stack AI application to safe enterprise-style tools.
Deploy
Containerize, test and release into an authenticated cloud environment.
Measure
Evaluate quality, latency, reliability, operating cost and value.
Hand over
Deliver the runbook, user guide, adoption plan and executive demo.
Your 12-week journey
One application. Five learning phases.
Engineering foundations
Python, Git, AI-assisted development, APIs and data models.
LLM apps & interfaces
Provider adapters, structured outputs, streaming UI and approvals.
Knowledge & retrieval
Grounded RAG, hybrid search, reranking and orchestration.
Agentic workflows
Stateful agents, scoped tools, MCP and human review.
Production delivery
Evals, red-teaming, CI/CD, cloud deployment and capstone handover.
Delivery workshops run alongside technical work. The instructor confirms the detailed session calendar for each cohort.
12 technical modules
Python foundations for AI engineering
AI delivery and collaborative development
FastAPI, OpenAPI and PostgreSQL
LLM APIs and provider abstraction
React and human-in-the-loop interfaces
Grounded retrieval with RAG
LangChain retrieval orchestration
LangGraph agent workflows
Agents SDKs and MCP integrations
Evaluation, guardrails and red-teaming
Containers and continuous delivery
Cloud deployment and operations
Six client-delivery workshops
Discovery and business baseline · Scope and acceptance criteria · Architecture and integration design · Stakeholder demos and user acceptance · Launch and security review · Value measurement and handover
Enrolled learners receive the full 45-lab sequence, project execution guides, acceptance checks and troubleshooting notes.
The system you build
Every module adds a working layer.
Front end · M5React, Vite, TypeScript, streaming chat and human approvals
API layer · M1–M3Typed Python, FastAPI, OpenAPI contracts and PostgreSQL
Agents · M7–M9LangChain, LangGraph, Agents SDKs and MCP integrations
Model providers · M4OpenAI and Claude, with Bedrock, Azure OpenAI and Ollama options
Data & retrieval · M3 / M6PostgreSQL, pgvector, hybrid search, reranking and citations
Quality & security · M10LangSmith, RAGAS, DeepEval, Presidio and pytest
Ship & run · M11–M12Docker, GitHub Actions, AWS or Azure, logs and Grafana
Choose a client problem
Five enterprise project options.
Build one primary scenario throughout the course, then extend it into your capstone.
IT Helpdesk Copilot
Answer knowledge-base questions with citations and propose ticket actions.
Claims Triage Agent
Extract document fields and route incomplete claims to a human reviewer.
HR Onboarding Assistant
Answer policy questions and assemble joiner checklists.
Invoice Reconciliation
Compare invoices with purchase orders through an ERP-style API.
Sales Account Intelligence
Create a source-cited account brief from CRM and permitted research data.
Bring your own problem
Propose a workflow from your experience. An instructor reviews the scope, data permissions and feasibility before you build.
Training uses synthetic, permissioned or de-identified data and mock enterprise integrations. Claims, HR and invoice scenarios require human review; exercises do not automate eligibility, employment or payment decisions.
Your portfolio evidence
A deployed system and the work behind it.
Live app & private repository
An authenticated deployment with CI history, tests and setup documentation.
Architecture pack
C4 diagrams, integration boundaries and ADRs explaining your choices.
Evaluation report
Groundedness, relevance, tool accuracy, red-team findings and fixes.
Client artifacts
A problem brief, PRD-lite, SOW outline and acceptance criteria.
Handover pack
Runbook, user guide, rollout risks and adoption plan.
Executive demo
A ten-minute presentation of the workflow, measured results and remaining limits.
Results are measured on your own assessed training project. No example metric is presented as a guaranteed Neriya learner outcome.
Who this is for
Bring your coding foundations. Build applied AI skills.
Designed for software developers, data and ML engineers, consultants, technical leads and enterprise application specialists. Prior AI or machine-learning experience is not required.
What you should already know
- One programming language, ideally Python
- Command-line basics, Git and pull requests
- How APIs and relational databases work
- Time for live sessions plus 3–4 hours of weekly self-practice
What you need
- A laptop with at least 8 GB RAM; 16 GB recommended
- Permission to install Docker, Node.js, Python and Git
- GitHub and approved model-provider access
- An AWS or Azure practice account with spending limits
Provider API usage, cloud consumption and separately quoted licenses may cost extra. Confirm access, costs and cohort times with enrollment support before payment.
Practice with purpose
Six priorities throughout the program.
Build first
Use guided labs and build sprints to turn concepts into working features.
Compare providers
Evaluate OpenAI and Claude through a common adapter and measurable tradeoffs.
Connect the whole stack
Move from React and FastAPI through retrieval to a deployed application.
Test for production demands
Practice evaluation gates, guardrails, access controls and observability.
Work with stakeholders
Rehearse discovery, scope decisions, progress demos and handover.
Explain your evidence
Use 17 assessed deliverables to discuss behavior, failures and design choices.
Career directions
Skills for client-facing AI delivery.
Forward Deployed Engineer
Work with customer teams to integrate and deliver AI in their environment.
Applied AI Engineer
Build and evaluate LLM, retrieval and agent features in products.
AI Solutions Engineer
Design and demonstrate solutions against customer requirements.
Agentic AI Developer
Build supervised workflows across enterprise tools and data sources.
Use your project to explain a business problem, demonstrate a working system and defend your engineering decisions. Resume reviews, interview coaching and any additional career services must be confirmed with enrollment support.
Job requirements vary by employer and seniority. Course completion does not guarantee employment, salary, an interview or affiliation with any named technology company.
INVEST IN YOUR PRACTICAL SKILLS
Tuition & enrollment
$3,900 USDTotal course tuition · 12 weeks. Independent certification exams and separately quoted third-party licenses are excluded. Review the cohort schedule and payment terms before enrolling.
Before you enroll
Your questions, answered.
Who is this program for?
Developers, data/ML engineers, technical consultants and enterprise application specialists who already understand a programming language, basic Git, APIs and databases.
How much time should I plan for?
The program spans 12 weeks. Confirm live-session days and hours for your cohort before enrollment, and plan for 3–4 hours of weekly self-practice in addition to live sessions.
Is the training live or recorded?
The course includes live instruction. Assigned classroom links, cohort schedules and published recordings remain in the existing enrolled learning room. Confirm recording availability and access terms with enrollment support.
Are there 12 modules or 18?
There are 12 technical modules containing 45 labs, plus six Neriya client-delivery workshops: 18 learning units in total. They lead to 17 assessed checkpoints and one integrated capstone.
What will I have at the end?
On completing the assessed work, you will have a deployed capstone, a private code repository, architecture decisions, evaluation/security evidence, client artifacts, a handover pack and an executive demo.
Can I pay in installments?
Use Review installment plan to see the available payment amounts, dates and terms before you agree to automatic payments. Full payment and installments apply to the same course.
Do you guarantee a job or salary?
No. Outcomes depend on your previous experience, the quality of your work, employer requirements and the market. Review Neriya’s existing enrollment and payment policies for applicable terms.
Can my employer sponsor me or a team?
Contact enrollment support to confirm invoicing, funding documentation, cohort arrangements and any team proposal. Sponsorship and third-party funding approval are determined separately.
Where are my detailed labs and course sessions?
Sign into your verified learner account with active enrollment. Open my full enrolled curriculum on this page for the lab and project guide; use Open my enrolled course for assigned live sessions and recordings.
Take AI from idea to deployment.
Build a system you can demonstrate, measure and explain.
Explore enrollment →Your enrolled learning room
Open assigned live sessions, classroom links and published recordings using your existing learner account.