6 weeks · 12 sessions · 36 live hours
AI Automation & Agents Bootcamp
A lab-heavy bootcamp for the people who will own AI systems: orchestration, retrieval, evaluations, monitoring and cost control.
Overview
Design, build and run AI agents and automations in production.
Who it's for: Operations, product and technical teams, developers and analysts.
Prerequisites
Comfort reading JSON and calling an API, or having automated something before in n8n, Zapier or a scripting language. You will need accounts with an AI provider and a place to deploy.
Assessment
Three graded builds during the program, plus a capstone agent reviewed against a rubric covering reliability, evaluation, guardrails and cost.
Certification
Indus AI Academy certificate, issued by INDUS AI Private Limited, awarded on three reviewed builds and a production-ready capstone agent.
- Duration
- 6 weeks · 12 sessions · 36 live hours
- Format
- Live online cohort · 2 sessions a week with build labs
- Level
- Intermediate · basic scripting helpful
- For
- Operations, product and technical teams, developers and analysts
Next cohort
Cohort 01 · starts 16 November 2026
Mondays and Wednesdays, 7–10 PM IST
Ends 23 December 2026 · 20 seats
₹35,999₹44,999per seat + GST
Early bird: first 10 seats, booked by 5 November 2026.
Private cohort for your company from ₹4,50,000.
Team pricing available for four or more seats.
You will leave with
- Three production-grade agents, built and reviewed
- Patterns for RAG, tool use and multi-agent workflows
- Evaluation and monitoring set-ups you can reuse
- Deployment and cost-control playbooks
Curriculum
Session by session: AI Automation & Agents Bootcamp
12 sessions · 36 live hours
- 013 hrs
Agent architectures
You build: A working agent loop with two tools
- Model, tools, instructions, memory
- When an agent beats a fixed workflow
- Failure modes to design around
Practice: Re-run your agent on ten inputs and record every failure.
- 023 hrs
Tool design and APIs
You build: Three reliable tools with clean schemas
- Designing tool inputs a model can get right
- Errors, timeouts and retries
- Rate limits and idempotency
- 033 hrs
Retrieval over your own knowledge
You build: A retrieval layer over real company documents
- Chunking, embeddings and hybrid search
- Citations and grounding
- Keeping the index fresh
Practice: Measure answer quality before and after retrieval tuning.
- 043 hrs
Memory and state
You build: An agent that holds context across a conversation
- Short-term versus long-term memory
- Threads, sessions and user state
- Forgetting on purpose, and privacy
- 053 hrs
Orchestration and multi-agent workflows
You build: A two-agent workflow with routing
- Routing, planning and hand-offs
- When multi-agent helps and when it hurts
- Queues and long-running work
- 063 hrs
Evaluations
You build: An evaluation harness with a golden set
- Building a golden set from real cases
- Model-as-judge, and its limits
- Regression testing before each change
Practice: Add twenty cases to your golden set and run a regression.
- 073 hrs
Guardrails and safety
You build: A guardrail layer around your agent
- PII detection and redaction
- Prompt injection and untrusted content
- Approval gates for risky actions
- 083 hrs
Voice agents
You build: A working voice agent prototype
- Speech to text and text to speech pipelines
- Latency budgets for natural conversation
- Indian languages, accents and code-switching
- 093 hrs
Integrations with real systems
You build: An agent connected to a CRM or helpdesk
- Webhooks, queues and back-pressure
- Auth, secrets and least privilege
- Writing back safely
- 103 hrs
Deployment
You build: Your agent deployed and reachable
- Hosting choices and environments
- Secrets, config and rollbacks
- Cost control and budgets
- 113 hrs
Monitoring and iteration
You build: Tracing, alerts and a triage runbook
- Tracing every step of a run
- Triaging failures from real traffic
- Feedback loops that improve the agent
- 123 hrs
Capstone review
You build: Your capstone agent, reviewed against the rubric
- Reliability and evaluation review
- Cost and latency review
- Hand-over documentation
Certification
A certificate backed by a registered company and real work.
Indus AI Academy certificates are issued by our parent company, and only once your capstone project has been reviewed, so they stand for skills you can demonstrate.
- Issued by INDUS AI Private Limited
- DPIIT-recognized startup (Startup India)
- ISO 9001:2015 certified
- MSME registered
- Awarded only after a reviewed capstone project
- Ready to share on LinkedIn and your CV
Certificate of completion
This certifies that
Your Name
has completed the AI Automation & Agents Bootcamp, including a capstone project reviewed by faculty.
Vivek Gupta
Founder · IIT Delhi alumnus
INDUS AI Private Limited
Issuing company
FAQ
Questions about AI Automation & Agents Bootcamp
What background do I need for the AI Automation & Agents Bootcamp?
Basic comfort with scripting helps. The bootcamp is built for operations, product and technical teams, developers and analysts who will own AI systems.
What will I have built by the end of the bootcamp?
Three production-grade agents, built and reviewed, plus reusable patterns for retrieval, evaluation, monitoring, deployment and cost control.
Which tools and models does it use?
Provider-neutral. We work with the major hosted models and open alternatives, orchestration in n8n and code, and the observability tools teams actually run.
More programs
Explore other AI courses.
AI Generalist Certification
Become the person on your team who gets things done with AI.
View program6 sessions · 12 live hours, or 2 days in personAI for Leaders
Set direction, back the right bets and lead an AI-ready organization.
View program1 day · 4 sessions · 6 live hoursAI Foundations Workshop
A one-day, hands-on start for anyone new to AI at work.
View programReady to join AI Automation & Agents Bootcamp?
Book your seat in the next cohort, or set up a private batch for your company.