18-Week Executive ProgramApplication-Based Cohorts
Generative AI &Agentic AIEngineer.
Build it. Ship it. Own the outcome.
Build LLM applications, RAG systems and multi-agent workflows, then secure, evaluate and deploy them on real enterprise stacks—with Deepesh Mishra.
Two live sessions weekly · Hands-on projects · 1:1 mentorship
Application-based admission
Build GenAI systems teams can ship.
Share your details to discuss fit, cohort dates and investment.
Live architecture sessions.
Weekly labs. Demo Day.
See the program
A closer look at the engineer you can become.
Hear the story behind the program and how the learning connects engineering craft with real deployment work.
Talk with admissionsFrom prompt to production
Engineers who can
own the full system.
Move from model calls to dependable products: connect data, retrieve the right context, give agents governed tools, measure what works and ship safely.
The engineering gap
From AI demos to systems
teams can trust.
Teams need engineers who can build LLM applications, integrate models with real data and tools, and stay accountable for quality, security and operation.
Builder
Builds LLM applications, RAG pipelines and agents that work on real, messy data.
Integrator
Connects models to databases, APIs, Jira, Slack and cloud so AI fits real workflows.
Owner
Evaluates, secures and operates AI in production, staying responsible for the outcome.
A career-focused program, not a placement guarantee. Hiring outcomes depend on your effort and the market.
Career directions
Build toward the work
you want to do.
Choose a path where your existing engineering, data, QA, operations or domain experience can meet applied AI delivery.
GenAI / LLM Application Engineer
RAG and Knowledge Engineer
AI Agent Developer
AI Solutions / Forward Deployed Engineer
LLMOps / AI Platform Engineer
AI Security and Governance Engineer
Project studio
Four client-style projects.
No toy apps.
Each project has a client brief, stakeholders, constraints, security requirements and a panel review. The deliverables are deployed, documented and portfolio-ready.
Knowledge Copilot
Hybrid RAG, secure text-to-SQL, role-based access, PII masking and guardrails. Deliver discovery notes, architecture and a stakeholder demo.
Python · FastAPI · PostgreSQL · DockerMulti-Agent Ops Copilot
LangGraph supervisor, MCP tools, Jira and ServiceNow integrations, human approval and a cost-and-trace dashboard.
LangGraph · MCP · Jira · Slack · KafkaProduction Hardening
Security review, compliance pack, red-team report, UAT, runbooks and client handoff for a regulated deployment.
Okta · OpenTelemetry · Grafana · TerraformYour domain. Your client.
Take a full project from discovery to handoff, then defend your decisions at Demo Day.
Databricks · Snowflake · CI/CD · KubernetesThe roadmap · 18 weeks
Five phases. Four builds.
One engineer who can ship.
Each week pairs live architecture sessions with a lab, a deliverable and tools used by production AI teams. Open a phase to see the details.
01Weeks 1-3Foundations
Understand LLMs. Build like an engineer.LLM fundamentals, prompt design, structured outputs and deployable GenAI APIs.
- Transformers, tokens, embeddings and context
- Compare OpenAI, Claude, Gemini and open models
- Structured JSON, Pydantic, tool calling and eval-first prompts
- Async Python, FastAPI, auth, retries, caching and Docker
Hands-on labCompare three model providers, then build a tested GenAI API.
You shipPrompt and evaluation kit · tested API service
02Weeks 4-7RAG systems
Build AI that survives real data.Retrieval, knowledge graphs, multimodal documents and measurable answer quality.
- Embeddings, chunking, metadata and vector stores
- Hybrid search, reranking, query rewriting and multi-hop retrieval
- Knowledge graphs and GraphRAG with Neo4j
- Multimodal documents, OCR-free parsing and secure text-to-SQL
Hands-on labBuild and evaluate RAG over messy documents, tables and business data.
You shipProject 1 · Knowledge Copilot with hybrid RAG, RBAC and stakeholder demo
03Weeks 8-12Agentic AI
Systems that plan, act and connect safely.Stateful agents, MCP, multi-agent orchestration, approvals and enterprise integrations.
- ReAct, plan-and-execute, routers and LangGraph state
- Tool registries, MCP servers, scopes and audit trails
- Supervisor, parallel and hierarchical multi-agent patterns
- Human approval, memory, agent evaluation, Jira, Slack and ServiceNow
Hands-on labBuild a supervisor-based operations copilot with governed tools and approvals.
You shipProject 2 · Multi-Agent Ops Copilot with traces, integrations and cost dashboard
04Weeks 13-16Production & security
Ship systems enterprises can approve.Guardrails, identity, observability, cost engineering, cloud deployment and governance.
- Prompt injection, PII masking, OAuth, SSO and RBAC
- Threat models, red-teaming and blast-radius controls
- LLMOps, tracing, CI evaluation gates, canaries and rollback
- Docker, Terraform, cloud deployment, SLOs and compliance essentials
Hands-on labRed-team the system, harden it, deploy it and run an incident drill.
You shipProject 3 · Production hardening, compliance pack, UAT and handoff
05Weeks 17-18Capstone & Demo Day
Run the full lifecycle. Defend your decisions.Choose BFSI, healthcare, enterprise operations or a domain of your choice.
- Discovery, scoping and architecture decision records
- Build with RAG, agents and integrations
- Add evaluation, security controls and CI/CD
- Present to an industry-style review panel and prepare for interview loops
Hands-on labWeekly mentor reviews, Demo Day, two mock interview loops.
You shipDeployed capstone · portfolio · interview and job-search plan
Portfolio · Resume · LinkedIn · Interview loops · Referrals
The technology
Learn the trade-offs.
Not just the tool names.
Work across models, orchestration, retrieval, integrations and production operations. Tool choices follow the job each system needs to do.
What you can do by the end
Evidence you can explain.
Systems you can defend.
Build a practical portfolio around the capabilities the brochure targets, without promising a particular hiring result.
Build reliable LLM applications with prompts, structured outputs and tools.
Design secure RAG, graph and multimodal systems over messy data.
Connect agents to databases, Jira, Slack and enterprise APIs with human approval.
Measure quality, cost and latency with repeatable evaluation.
Defend systems against prompt injection, data leakage and excessive agency.
Deploy on AWS, Azure or GCP using Docker, Terraform and CI/CD.
Explain and defend engineering decisions in GenAI and Agentic AI interviews.
The learning environment
Work in the room.
Ship work that travels.
Live cohorts, hands-on labs, client-style reviews and a final capstone defense make the work visible from week one.
Career launch · every week
Career preparation built alongside your engineering work.
Develop a portfolio and interview story from the systems you actually shipped—not a separate set of generic career exercises.
Ask about admissions- Role targeting and a skills-gap audit
- GitHub portfolio setup and project documentation
- GenAI and Agentic AI resume, LinkedIn and Naukri reviews
- Technical and system-design interview question bank
- Two 1:1 mock interview loops with feedback
- Referral and outreach playbook
No placement, salary or hike is guaranteed. Admissions shares the current cohort dates, fees, instalments and scholarship details.
Your lead mentor
Deepesh Mishra
Founder, Datamarcos · Solution Architect · Corporate Trainer · M.Tech, BITS Pilani
18+ years of industry experience and 25,000+ professionals trained across 200+ organisations in 10+ countries. Deepesh teaches the way engineers work: scope first, then build, secure, deploy and hand over.
Learner feedback
What learners say about learning with Deepesh.
Feedback shown here comes from the brochure and spans Datamarcos data and AI learning programs.
“His deep understanding of the subject, combined with a clear and engaging teaching style, made complex concepts easy to grasp, even for those new to the field.
“Learning under Deepesh Sir has been a great opportunity. All my doubts were dispelled and I was able to hone my skills with great hands-on activities and projects.
“Many thanks for the great Databricks training at ZS! Let's stay connected.
Application-based admissions
A focused cohort, selected for the work.
Move through a clear fit process before cohort allocation. The goal is to make sure the pace and practical format suit your starting point.
Apply
Share your background and the role you are working toward.
Diagnostic
Technical and fit assessment: Python, APIs and problem-solving.
Interview
Discuss your goals with an admissions mentor.
Offer
Cohort allocation, bridge pre-work and onboarding.
Bridge pre-work in Python, Linux, Git and SQL is provided before Day 1.
The full program guide
Review the brochure.
FAQ
Before you apply.
Program details grounded in the current brochure. Admissions can confirm cohort-specific information.
01Do I need prior AI experience?Read answer
No. The brochure expects Python basics and working engineering fundamentals or strong shipped projects. Bridge pre-work in Python, Linux, Git and SQL is provided before Day 1.
02How much time should I plan for each week?Read answer
Plan for about 8-12 hours per week, including two live sessions, a weekday lab and project work.
03How is the program delivered?Read answer
It is an 18-week executive program with two live sessions each week, weekly delivery labs, project studio, scheduled 1:1 mentorship and a final Demo Day.
04What projects will I build?Read answer
Three client-style projects: a secure Knowledge Copilot, a Multi-Agent Ops Copilot and a Production Hardening project, followed by a domain-specific capstone.
05Will I get a job after the program?Read answer
No placement, salary or hike is guaranteed. The program includes deployed projects, portfolio support, interview preparation, mock loops and a referral/outreach playbook.
06What is the fee and next cohort date?Read answer
Admissions shares current cohort dates, investment, instalment options and scholarship details after a fit conversation.
07Can my company sponsor me or join as a team?Read answer
Yes. Contact admissions to discuss corporate sponsorship and team cohorts.
Generative AI & Agentic AI Engineer
Become the engineer every AI team wants.
Request a conversation about fit, cohort dates, fees, instalments and scholarship options.
Download the brochure firstThe program does not promise placements, salaries or hikes.