You Understand the AI. Now Build the Proof.
Two Working AI Projects in 6 Guided Weeks — and the Language to Defend Every Decision in Them.
You already know how retrieval and agents work. What you are missing is a repository someone else can open, read and believe — and a clear account of why you built it that way. Where this takes you: two finished systems, one RAG application and one agentic workflow, with architecture diagrams, READMEs and decision notes in your own words. The program is live right now, so you start the moment you enrol — six guided modules, with live sessions continuing on Thursdays. Stage 4 of the Manifold ladder, where proof of work lives. Systems ship. Demos don't.
- ✓ 2 Practical AI Projects
- ✓ 2 GitHub Repositories
- ✓ 2 Architecture Diagrams
- ✓ 2 Professional READMEs
- ✓ Resume + LinkedIn Positioning
- ✓ Project Explanation Frameworks
You build it. We direct it. You leave able to walk anyone through it.
You've Learned AI.
But What Can You Actually Show?
Courses Completed. Portfolio Empty.
You've watched the videos and understood the concepts — but your GitHub still doesn't reflect your AI learning.
Another Chatbot Won't Differentiate You.
Most beginner projects look identical. You need projects built around a real use case and a clear architecture.
You Built It. But Can't Explain It.
The code runs, but when someone asks why you chose the architecture, your explanation becomes difficult.
Half-Finished Projects Everywhere.
Repositories, notebooks, experiments — but nothing packaged into something you feel confident sharing.
You Know RAG and Agents in Theory.
But you don't yet have tangible proof that connects those concepts to something you built.
Knowledge becomes far more valuable when you turn it into proof.
Two Projects.
Built Properly. Explained Clearly.
Ten shallow repos prove nothing. You build two systems instead — deep enough that you can open either one in front of a panel and walk through the use case, the architecture, the trade-offs you accepted, and what you would change next.
You won't just leave with code.
You'll leave knowing how to talk about what you built.
By Week 6, You Won't Just Say "I'm Learning AI."
You'll Have Proof.
RAG / Knowledge AI Application
- ✓ Working application
- ✓ GitHub repository
- ✓ Architecture diagram
- ✓ Professional README
- ✓ Project explanation
- ✓ Resume bullet
- ✓ LinkedIn project story
- ✓ Technical decisions
Agentic AI Workflow Application
- ✓ Working agentic workflow
- ✓ GitHub repository
- ✓ Architecture diagram
- ✓ Professional README
- ✓ Project explanation
- ✓ Resume bullet
- ✓ LinkedIn project story
- ✓ Technical decisions
Two projects. Two stories.
One much stronger AI portfolio.
Build a Practical RAG / Knowledge AI Application
Recommended default: an AI Knowledge Assistant or Document Intelligence Assistant — grounded in a use case that actually means something on your resume.
Use Cases You Can Choose
- Internal knowledge assistant
- Customer support knowledge assistant
- Policy / documentation assistant
- Research assistant
- Domain-specific knowledge application
What You'll Understand & Build
- Business problem & target user
- Document / data ingestion
- Chunking strategy
- Embeddings & retrieval
- Prompt + retrieved context
- Answer generation
- Grounding & source references
- Simple application flow (UI or API)
- Clean repository structure
Build a Practical Agentic AI Workflow
Choose from an AI Research Assistant, a Customer Request Resolution Workflow, an AI Operations Assistant, or a Multi-Step Business Workflow Assistant.
What You'll Understand & Build
- User request handling
- Task classification & planning
- Model interaction
- One or two working tools
- Structured outputs
- Workflow & state thinking
And the Decision Layer
- Review / decision points
- Human-in-the-loop concept
- Final response & output shaping
- Clear, explainable architecture
- Repository structure that reads well
You Will Be Guided Through the Build.
Nobody hands you a title on Day 1 and wishes you luck. You make the design calls in a live session, with direction available at the point you would otherwise stall for a week.
6 Modules.
One Clear Portfolio Outcome.
Choose a Portfolio-Worthy AI Project
Turn a vague project idea into a structured, defensible use case.
Create your Project 1 brief.
Project Brief + Architecture Canvas
"I know exactly what I am building and why."
Build Project 1 — RAG / Knowledge AI Application
Your first working portfolio project, built end to end with guidance.
Build Project 1.
Working Project 1 + GitHub Repository
"I have my first working portfolio project."
Turn Project 1 Into a Professional Portfolio Asset
Move from "it runs" to "I can present this to anyone."
Package Project 1.
Portfolio Pack #1
"I can explain this project, not just run it."
Build Project 2 — Agentic AI Workflow
A second project that shows a different class of AI capability.
Build Project 2.
Working Agentic AI Project + GitHub Repository
"I now have a second project that demonstrates a different AI capability."
Explain Architecture, Decisions & Trade-Offs
The week that turns a project you built into a project you can discuss with confidence.
- What if usage increases?
- What if the model fails?
- What if the tool fails?
- Where should secrets live?
- What would need authentication?
- How would quality be measured?
- What should be logged?
- What should eventually be monitored?
You'll learn the next questions experienced engineers ask — the ones that turn a good project conversation into a senior one.
Document your decisions and trade-offs.
Technical Trade-Off Notes + Portfolio Pack #2
"I can discuss why I built it this way."
Package Your AI Portfolio Like a Professional
Everything you've built becomes something you can send in a single link.
Repository naming · README structure · screenshots · architecture · setup · project explanation
Strong project bullets · architecture + responsibility + outcome · no buzzword stuffing
Business problem · solution · architecture · learning · project link
Assemble your complete portfolio.
Final AI Portfolio Kit
"I now have two AI projects I can confidently show and explain."
Six Weeks Later,
Here Is What You Can Put in Front of Someone.
This is not another folder of course notes.
This is your AI proof-of-work portfolio.
What Makes These Projects Different
From Another Tutorial?
Typical Tutorial
- — Watch
- — Copy
- — Run
- — Forget
This Accelerator
- ✓ Understand the problem
- ✓ Design the architecture
- ✓ Build
- ✓ Personalize
- ✓ Explain decisions
- ✓ Package
- ✓ Showcase
The goal isn't more code.
The goal is stronger proof of work.
Hands-On Enough to Build.
Clear Enough to Explain.
By the end you can do two things that usually come apart: build the thing, and account for it. Architecture, technical reasoning, your own scope of responsibility, and the packaging that makes all of it legible to someone who was not there.
- "What did you build?"
- "Why did you design it this way?"
- "What was your responsibility?"
- "What would you improve next?"
You do not need to become an expert in every infrastructure layer before building a portfolio. You need to understand your project well enough to own the conversation about it — and that is exactly what these six weeks are built around.
Build the Project — and learn what comes next.
You'll know what questions come next — without losing focus on completing your portfolio.
Taught Live by Nachiketh Murthy
Nachiketh has taught 100,000+ engineers across online courses, YouTube, and live cohorts — including senior engineers and engineering leaders from companies like Micron and Salesforce who join his live bootcamps. He builds and teaches production Agentic AI systems: LangGraph orchestration, RAG grounding, evaluation, observability, and AWS/Azure/GCP deployment. His teaching style is hands-on and architecture-first — every concept lands in code you ship, and every design decision comes with the trade-off reasoning behind it.
He runs these sessions live himself. You are not watching a recording of someone else’s cohort — you get every module released so far plus the live guidance sessions still ahead, where you can ask him why a decision was made while it is being made.
Engineers Who Have Built
in Nachiketh’s Live Cohorts.
The AI Project Portfolio Accelerator is a new program. The experiences below are from engineers who previously attended Nachiketh’s live Manifold AI Learning cohorts and reflect the teaching quality, hands-on depth and live learning experience — not this exact curriculum.
“I wanted a curated course on how we actually build solutions and make them production-level rather than only building POCs. Enterprise RAG, observability, prompt versioning — I can explain all of it better now.”
“The weekend bootcamps are well-structured, combining concepts with hands-on experiential learning — a strong blend of theory, practical implementation, and real-world lessons.”
“An amazing program — it covered every aspect of a production project, from requirements to testing to final deployment.”
“I’m a backend Java engineer. With no prior exposure to AI, I got a good solid foundation and a clear direction. This program changed my thinking about how we should implement enterprise-level RAG and build production-ready agents.”
“Before joining, I struggled with GenAI concepts. The cohort helped me bridge the gap between a Data Scientist role and a GenAI role. I would highly recommend it to anyone who wants to transition with a strong foundation.”
9.6/10 average experience rating in our latest cohort feedback survey.
Built for Engineers Who Already Ship
and Have Nothing to Point At.
✓ This is for you if…
- ✓ You have learned AI concepts but lack strong projects.
- ✓ Your GitHub does not reflect your AI learning.
- ✓ You want practical RAG and Agentic AI projects.
- ✓ You struggle to explain project architecture.
- ✓ You want to turn projects into resume and LinkedIn assets.
- ✓ You prefer guided live building over solo guesswork.
- ✓ You want accountability to actually finish.
- ✓ You have basic Python familiarity.
— Probably not the right fit if…
- — You have absolutely zero programming exposure.
- — You are looking only for theoretical lectures.
- — You expect somebody else to complete your projects.
- — You are unwilling to work on weekly project tasks.
Typical profiles: Software Engineers · Backend Engineers · QA & Automation Engineers · Data Engineers · Data Scientists · Cloud Engineers · DevOps · MLOps Engineers · Technical Leads · Architects moving toward AI
6 Modules. Live Guidance.
2 Finished Projects.
Where This Sits
on Your Path.
- ✓ 6 live guided sessions
- ✓ 12 hours live
- ✓ 2 practical AI portfolio projects
- ✓ Architecture guidance
- ✓ Project templates
- ✓ README frameworks
- ✓ GitHub packaging guidance
- ✓ Resume project positioning
- ✓ LinkedIn project positioning
- ✓ Project explanation frameworks
- ✓ Technical trade-off framework
- ✓ Final portfolio packaging
- ✓ Session recordings — lifetime access
What Engineers
Ask Before Joining.
What exactly will I have after six weeks?
Two completed AI projects — one RAG / Knowledge AI application and one Agentic AI workflow — along with GitHub repositories, architecture diagrams, READMEs, resume bullets, LinkedIn positioning, and project explanation material for each.
Is this just another coding course?
No. Coding is part of the experience, but the accelerator also teaches you how to structure, understand, explain, and package your projects professionally — which is where most learners actually get stuck.
Will I actually build during the accelerator?
Yes. The accelerator is designed around guided building and weekly implementation. You will be guided, but you will build.
How many projects will I build?
Two. The focus is depth, completion, and clarity rather than ten superficial projects you'd struggle to explain.
What kind of projects?
One practical RAG / Knowledge AI application and one practical Agentic AI workflow — two genuinely different AI capabilities on your portfolio.
Do I need Python?
Basic Python familiarity is recommended. You should be comfortable reading and modifying simple Python code.
Can I put the projects on GitHub and my resume?
Yes, after completing and personalizing your projects. The accelerator specifically includes portfolio packaging for GitHub, resume, and LinkedIn.
Will this help me explain projects in interviews?
Yes, naturally. You'll learn how to explain the problem, architecture, your responsibility, decisions, challenges, and future improvements — the same structure that makes any project conversation stronger.
Do I have to do work between sessions?
Yes. Some weekly project work is expected. The live sessions provide direction and guidance; completing and personalizing the project is part of the learner journey — and it's what makes the final portfolio genuinely yours.
How deep do we go technically with the projects?
You'll build practical end-to-end portfolio applications and understand their architecture, technical decisions, and improvement paths. We'll also introduce the questions experienced engineers consider as these systems grow — reliability, quality, security, monitoring, cost, and scale.
Are recordings available?
Yes — every session is recorded and you get lifetime access to all recordings. Start today with everything already released, rewatch any module any time, and catch up on your own schedule without falling behind.
Does this guarantee a job?
No. The accelerator strengthens your proof of work and your ability to present your AI projects professionally. It does not provide job guarantees.
Stop Collecting Tutorials.
Start Building Proof.
Six guided weeks from here: two systems that run, two repositories someone can read, and a clear account of every decision inside them.