Fellowship Program in AI Engineering
Become an AI Engineer who ships real AI apps — not just notebooks.
Fellowship Program in AI Engineering
Become an AI Engineer who ships real AI apps — not just notebooks.
6 Guided Production Projects — Plus Your Own Capstone
Click an app to see what you'll ship. Each one is deployed, evaluated and monitored — with a one-page spec, success metrics and a demo.
Billing-Policy.md Refunds-FAQ
AskDocs
Cited Q&A over a real doc source (Notion / Google Drive) with hybrid retrieval + reranking, conversation memory and guardrails — plus an agentic-RAG extension. Evaluated with RAGAS, traced in Langfuse.
Sample screens — you build, deploy and operate each one yourself.
Who This Program Is For
Built for software engineers moving into AI. No prior ML or GenAI experience needed.
Grads with substantial project work
You've built and shipped large software projects. Add the AI-app layer — RAG, agents, evals, deployment — and graduate with a portfolio most peers don't have.
Software, full-stack & backend engineers
Own AI systems end to end — retrieval, orchestration, cost/latency, observability — and lead the AI work on your team.
Senior engineers, leads & architects
Decide what to build and prove it works: specs, success metrics, evals, PoC → production, stakeholder demos.
ML engineers & data scientists
Move from notebooks and models to shipped products — agents, MCP, production serving, LLMOps.
Pre-requisites
- Python proficiency (functions, classes, typing, packaging)
- Git & GitHub — branching, PRs, basic CI
- HTTP/REST & JSON — calling and building a simple API
- SQL basics — joins, aggregates, relational schemas
- Command line & Docker basics
- Engineering-undergrad math intuition (linear algebra, probability)
Pre-work (~8–10 hrs, self-paced)
- Lab-stack setup — Docker, Python, Node; bring up the standard lab stack; API keys via the program gateway
- Python-for-production-AI refresher (typing/Pydantic, async, packaging)
- ML concepts primer (bias/variance, metrics, model families)
No prior ML, deep-learning or LLM experience required — the refreshers are built in.
Not sure you're a fit? Request a callback — a program advisor will tell you honestly whether to start now or prep first. No payment, no commitment.
Why AI Engineering, Why Now
Every product team is adding LLM features. Very few engineers can ship them reliably. The gap isn't prompting — it's retrieval, evals, agents, cost control and production operations.
What's the difference — and which one is this?
An AI Engineer builds and operates AI systems: retrieval, agents, evals, fine-tuning, serving. A Forward-Deployed Engineer takes those systems into a customer's organisation: discovery, scoping, integration, adoption — and owning the outcome.
This program is both, in that order. Six sprints build the AI-engineering core in depth. The problem → production (FDE) work cycle is layered on top — every guided project ships with a spec, success metrics and a demo narrative, and the capstone runs the full cycle — and a 2-week Interview Preparation Sprint targets AI Engineer and FDE interviews. You graduate able to go either way.
See problem → productionMaster AI Engineering With Key Industry Tools
25+ industry-relevant tools across LLM apps, agents, retrieval, evaluation, fine-tuning and production — taught concept-first, so the tools stay swappable as the ecosystem evolves.
Pythonlanguage
Git & GitHubworkflow · CI
Dockerpackaging
FastAPIserving
StreamlitUI
Rediscache · state
Postgres + pgvectorretrieval store
vLLMserving
PyTorchdeep learning
Hugging Facemodels · PEFT
ClaudeLLM
GeminiLLM
Ollamalocal models
LangGraphagents
LangChainorchestration
MCPtools · integration
Qdrantvector dbpgvectorretrieval
CrewAI / AutoGenintroduced
Grafanamonitoring
Prometheusmetrics
GitHub ActionsCI deploy gate
LiveKitvoiceClaude Codecoding agent
Cursorcoding agent
Kubernetesintroduced
Google Cloudintroduced
Notiondocs API
Google Drivedocs API
HubSpotCRM
Stripepayments · test modeGitHubrepos · PRs
Gitea · Plane · Mattermoststand-insModel hubpublishingBuilt On Four Pillars
The program follows the AI Engineering Skills Map published by DeepLearning.AI — with an FDE (Forward-Deployed) layer that most courses skip.
Building & deploying AI apps
LLM foundations, grounding with data (RAG), agentic systems, evaluation-driven development, operating in production, ML foundations.
Software-engineering fundamentals
System design, data, security & scaling — recapped for a software audience and deepened via LLD/HLD electives.
Using coding agents
Plan→execute→deploy workflow, autonomy & safe permissions, md-files/hooks/MCP, agentic code review, large-codebase work.
Shaping the build
Discovery & scoping, spec & success metrics, PoC→MVP→production, demo storytelling, owning outcomes.
The Curriculum To Crack AI Engineer Roles
Six 4-week sprints followed by a 2-week Interview Preparation Sprint. Each sprint: 10 live sessions (2 hrs) + ~20 hrs of take-home builds, ending in a demo day. Click a sprint to see what you'll learn and ship.
Foundations & the AI-App Mindset
Learn how AI applications differ from traditional software and master the LLM building blocks — with a hands-on ML/DL refresher.
- How AI apps differ — non-determinism and the build-examine-decide loop
- LLM foundations: tokenization, context windows, sampling, caching, tool calling, multimodal inputs
- Prompt & context engineering; structured (JSON) outputs and validation
- Design patterns for AI apps — and when not to use AI/agents
- Evaluation 101: the error-analysis loop and a first eval harness
- ML/DL refresher (bias/variance, metrics, transformers) and a computer-vision primer
- System-design recap for AI apps: API design, data models, auth, async
Sprint contents and project sequencing may be refined with industry experts before the cohort starts.
Data Structures & Algorithms
- DSA-101 · Beginner DSA4 weeks
- DSA-102 · Problem-solving foundations4 weeks
- DSA-201 · Intermediate DSA4 weeks
- DSA-202 · Trees, graphs & recursion4 weeks
- DSA-203 · Advanced DSA & DP4 weeks
System Design
- LLD-201 · Low Level Design — Foundations4 weeks
- LLD-202 · Low Level Design — Advanced4 weeks
- LLD-203 · Machine coding & interview practice4 weeks
- HLD-301 · Architecture & high-level design choices4 weeks
- HLD-302 · Distributed-system scenarios4 weeks
- HLD-303 · Building for scale on the cloud4 weeks
Projects You'll Build
Ship 17 builds — 10 mini projects, 6 guided projects and a capstone — plus 4 optional self-builds. Every build ends in a deployed artifact and a measured result: an eval report, a cost/latency dashboard, a demo — not a notebook.
Small, shippable builds that teach the muscle — each one ends in a working, deployed artifact and a measured result.
PromptForge
Prompt & context engStructured-output extractor / prompt playground with schema validation.
ModelSwitch
Routing · Evals 101Multi-model router with a first eval harness measuring tokens, latency, cost and quality.
SnapClassify
CV primerTransfer-learning image classifier deployed as a web app.
FindIt
Hybrid searchDense + sparse search fused with RRF and a reranker over a real dataset.
ToolAgent
Agents 101A ReAct agent calling a live API.
GateKeeper MCP
MCPA multi-tenant MCP server with RBAC, per-tenant scoping and audit logs.
EvalLab
EvaluationLLM-as-judge harness + a regression set wired in as a CI deploy gate.
TinyTune
Fine-tuningLoRA/QLoRA fine-tune + before/after eval + a distillation demo.
RouteCache
Cost & latencyModel-router + semantic-cache gateway; cost/latency before-vs-after.
SafeOps
Observability & guardrailsTracing + PII redaction + a blocked prompt-injection + a drift alert.
Learn To Decide What To Build — Not Just How
Forward-Deployed Engineers make AI work inside real organisations. We teach that work cycle — from problem framing to handover — and every guided project and the capstone practise it.
Frame the right problem
Discovery and scoping: find the problem worth solving with AI; define success metrics and write the one-page spec; MVP vs careful build.
Build within real constraints
Existing systems, data, security and tenancy; stack decisions — PoC → MVP → production.
Deploy, hand over & own
Eval gates in CI, fire-drills, runbooks and a handover doc; demoing to stakeholders; owning outcomes.
Harden for production
Multi-agent failure modes, RAG at scale, cost under load — the production-hardening clinic.
What Makes This Program Different
Most courses teach you about LLMs, RAG and agents. Here, you build, evaluate, deploy and operate them.
You Ship, You Don’t Watch
Every sprint ends in deployed artifacts — 16+ builds, not one capstone.
Concept-First, Stays Relevant
Context engineering, retrieval, the agent loop, harnesses, protocols and evals — taught above any specific model or tool, so it holds as tools evolve.
Production Skills, Taught & Graded
Evals as deploy gates, observability, retrieval engineering, MCP with RBAC/audit, cost/latency, security & compliance literacy.
Coding Agents As A Discipline
Three dedicated sessions plus a large-codebase project — workflow, safe autonomy, md-files/hooks, agentic review.
From Problem To Production
Problem framing, spec & success metrics, building within real constraints, deploying, handing over and owning outcomes.
Interview Preparation Built In
A 2-week sprint: system-design, decomposition and behavioural rounds — with practice sessions and scored feedback. Graduates can target AI Engineer, GenAI Engineer and Forward Deployed Engineer (FDE) roles.
What's Included In The Fee
The tools and cloud access you need for the projects are part of the program. We will use free tiers and resources wherever sufficient.
AWS Cloud Credits
ProvidedLLM API Access
ProvidedVector DB, Eval & Observability Tooling
ProvidedVoice & Deployment Tooling
ProvidedA Lab Kit for All 17 Builds
ProvidedStep-Up Reading Library
ProvidedEvery project comes with a starter repo, automated tests and a rubric. The local lab stack runs on your laptop with one Docker compose file, for free.
Learn From Industry Experts
Your mentors are shipping LLM, GenAI and agentic systems at product companies — teaching the live sessions and guiding your builds.
Soumya Mukherjee
Builds large-scale AI/ML platforms and brings LLMs and agentic AI into real production and automation workflows.
Deepak Sharma
Hands-on GenAI engineer who designs and ships RAG pipelines, agents and production LLM applications.
Career Outcomes & Support
A portfolio that reads like work experience — and career services, including 5 referral interviews, to turn it into the next role. Hover a step to see what it adds to your profile.
- Throughout · Sprints 1–6
A deployed portfolio
16+ live apps and a capstone, each with an eval report, cost/latency dashboard and demo video — hosted, linkable, reviewable.
- Final sprint · Interview Sprint
Resume & LinkedIn positioning
"Deployed on AWS", "built evals in CI", "shipped a fine-tuned model" — mentor-reviewed so it reads credibly to hiring managers.
- Weeks 25–26
Interview Preparation Sprint
2 weeks · 6 sessions + practice: agentic-system-design, ambiguous-problem decomposition and behavioural rounds; security & architecture interrogation; portfolio walkthrough — with scored feedback.
- On completion
Crio certificate
Certificate of completion from Crio.Do, backed by the graded builds behind it.
- After the program
Career services · 5 referral interviews
Five referral interviews for SDE or AI Engineer roles — plus resume & LinkedIn review, a portfolio walkthrough and scored interview practice.
Program Fee And Scholarships
One fee. No separate charges for cloud, API usage or tools — they're part of the program, and we use free tiers wherever sufficient.
Included in the fee
- 60+ live sessions (120+ hrs) with practitioners, plus the 2-week Interview Preparation Sprint
- Mentor-led capstone clinics and demo days
- Step-up reading library + optional self-paced modules (DSA, LLD/HLD)
- Crio certificate
- 17 builds (+4 optional self-builds) with starter repos and automated assessments
- Cloud credits, LLM API access and project tooling (free tiers wherever sufficient)
- Recordings of every live session
- Career services — 5 referral interviews for SDE or AI Engineer roles
What Crio Learners Say
Real experiences from developers who transformed their careers with Crio.
"The AI modules completely changed how I think about development. I went from writing code to architecting systems. The capstone project gave me a real AI product to showcase — it's what got me noticed at Flipkart."
"I was intimidated by LLMs, but the curriculum made it approachable. Hands-on with Claude, LangGraph, and building agents — not just watching videos. Three months after graduation, I landed a backend role that explicitly values AI skills."
"Through the Micro-Experiences I learned a bunch of industry-relevant skills. There were several modules where I was stuck, but the team at Crio always kept me motivated."
Get more insights on how Crio can help you become an AI Engineer
Grab our placement stats.
FAQs
Start Shipping Real AI Apps.
Apply now — a program advisor will walk you through fit, schedule and fees. No payment until you enrol.
- ApplyFill the form at the top — 30 seconds.
- Talk to a program advisorFit, schedule, fees and scholarships — and an honest read on whether this is the right program for you.
- Enrol and get your lab kitCloud sandbox, API keys and the Docker lab stack arrive before Day 1.
20, 100 Feet Rd, Vivek Nagar,
Chandra Reddy Layout, AVS Layout Ejipura,
Bengaluru, Karnataka 560095
