DEPARTMENT: AI ENGINEERING // GENERATIVE-AI-SYSTEMS-RAG-MULTI-AGENT-SWARMS
VIRTUAL INTERNSHIP45 Days (6 Weeks)•Department: AI Engineering
Generative AI Systems, RAG & Multi-Agent Swarms
A 45-day hands-on program on operationalizing large language models. Interns build low-latency Retrieval-Augmented Generation (RAG) pipelines, vector database embeddings, and autonomous multi-agent systems with tool-calling capabilities.
Tuition Fee₹299/ verified seat
Sprint Duration45 Days6 weekly phases
Lead MentorPriya NairSenior AI & LLM Systems Lead
MSME RECOGNIZED PROGRAMUDYAM-BR-28-0036675
Evaluated against Staff Engineer pull-request rubrics, automated CI code coverage thresholds, and immutable SHA-256 attestation.
SYLLABUS & MILESTONE DELIVERABLES
6 SPRINT PHASES // 45 Days (6 Weeks)Week-by-Week Technical Curriculum
WEEK 01GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100LLM Fundamentals & Structured Outputs
Core Engineering Topics:
Prompt engineeringTemperature & Top-PFunction calling
Required Task Deliverable:
Type-safe JSON generation pipeline with schema validation
WEEK 02GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100Vector Embeddings & Semantic Search
Core Engineering Topics:
Sentence transformersCosine similarityQdrant vector store
Required Task Deliverable:
High-speed semantic search engine across 10,000+ documents
WEEK 03GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100Hybrid RAG Architecture
Core Engineering Topics:
Recursive chunkingBM25 keyword + vector fusionCross-encoders
Required Task Deliverable:
Context-augmented Q&A system with cited sources
WEEK 04GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100Autonomous Agents & Tool Calling
Core Engineering Topics:
ReAct frameworkLangChain / LangGraphAPI tool execution
Required Task Deliverable:
Autonomous agent capable of querying APIs and generating reports
WEEK 05GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100Guardrails, Evaluation & Latency Optimization
Core Engineering Topics:
Prompt injection defensesRagas evaluation frameworkStreaming
Required Task Deliverable:
Hardened agent system with automated test evaluations
WEEK 06GATE_STATUS: MANDATORY EVALUATION
PR_MIN_PASSING: 75/100Capstone Multi-Agent Deployment
Core Engineering Topics:
Production deploymentTelemetry loggingFinal defense
Required Task Deliverable:
Deployed multi-agent enterprise software assistant
Competency Attestation Objectives
OUTCOME VERIFICATIONImplement semantic vector search with reciprocal rank fusion and sub-100ms retrieval latency
Construct hybrid RAG architectures with document chunking, metadata filtering, and re-ranking
Orchestrate multi-agent swarms with tool-calling capabilities and dynamic state management
Enforce strict hallucination guardrails and structured JSON output validation using Zod / Pydantic
Deploy scalable AI microservices with streaming token responses and WebSocket telemetry
CODE REVIEW RUBRIC
Senior Staff Review Matrix
Applied strictly to every pull request before sprint credit is granted.
RAG Retrieval Precision & Accuracy25 pts
Sub-100ms retrieval latency, high recall, and accurate citations
Agent Autonomy & Tool Reliability25 pts
Clean state management, zero infinite loops, and robust error recovery
Safety & Guardrail Compliance25 pts
Resistance to prompt injection and consistent schema adherence
System Architecture & Streaming UX25 pts
Fast time-to-first-token and clean streaming UI integration
COMPLETION CREDENTIALS
Certificate Delivery Options
Unlocked automatically upon achieving verified completion of all sprint tasks.
Option A: Verified E-Certificate₹149
Instant high-resolution PDF with verifiable cryptographic QR code and SHA-256 ledger digest.
Option B: Printed Hard Copy Certificate₹499
Premium textured parchment with holographic security seal dispatched to your India mailing address.
Ready to Begin Your Sprint?
Enroll today to receive your official offer letter and start working on Phase 01 tasks.