- Stochastic LLMs
- Chain-of-Thought Prompting and Few-shot Prompting
- Prompt Failure Modes
- Structured Outputs
- Prompting Reliably
- Prototypes and Demonstrations: 5
- System Design: Fact Checking Agentic System
A hands-on, no-fluff, and highly practical masterclass that helps you become great at designing, implementing, and shipping AI systems at your workplace.
Enrollments for Applied AI - Cohort 5 Cohort are open.
If you are looking for system design courses, I offer two - one for beginners and one for experienced engineers.
Prompt LLMs Reliably
Learn how to design reliable prompts, identify and prevent common LLM failures, and evaluate outputs systematically. This session establishes the foundation for building dependable AI applications.
Tool Use and RAG
Learn how production AI systems use tools and retrieval reliably at scale. This session covers the foundations of robust RAG systems, improving retrieval quality, reducing hallucinations, and handling real-world operational challenges.
Single Agent Loops and Systems
Learn the core patterns behind AI agents, how they reason, plan, and act across complex tasks. This session is essential for building reliable agents that can operate beyond single LLM calls.
Multi-Agent Systems and Memory
Learn how multiple AI agents collaborate, share memory, and coordinate effectively in complex systems. This session is essential for building scalable, maintainable agent architectures that remain reliable and observable in production.
Evaluation, Safety, and Systems
Learn how to evaluate AI systems systematically, measure performance, and identify weaknesses before deployment. This session is critical for building trustworthy, safe, and continuously improving AI applications in production.
Production Gotchas and Systems
Learn the architectural patterns behind scalable AI systems, from reliability and cost management to deployment and observability. This session is essential for building production-ready applications that perform consistently at scale.
I absolutely loved this cohort. The thing that stood out for me is the clarity of thought behind every reasoning and discussion. Things are highly practical and super intuitive.
I really loved how Arpit probed us during the AI system design discussion, which really lead to better design discussions.
I have built the multi agent systems that are currently working in production and i learned a lot from these sessions especially evals.
Thank you Arpit for everything!
I would strongly recommend this course to my colleagues who practically wants to build the AI systems.
... Every concept is grounded in real production experience, with practical insights into what you can expect when deploying agentic systems at scale ... More than specific techniques, you will learn strong mental models for designing and reasoning about these systems ... With so much noise around AI today, Arpit's cohort is a high signal, no-fluff learning experience ...
... it treats AI agents as distributed systems, not as a prompting exercise ... He always goes deep into technical topics, drawing on his own experience and first-principles thinking ...
... Before attending the course, I had many questions about how AI solutions are actually built in the real world ... This masterclass provided exactly that missing perspective. It helped me understand not only what can be built with AI, but more importantly how to think about designing AI-powered systems ...
... More than just the concepts, it was the way you broke down real-world systems, shared practical insights, and actually built prototypes instead of staying at the theory level ... The course didn't just teach me about agents—it changed the way I think about designing AI systems ...
You have some basic familiarity with Python
You know basic System Design
You have gone through, read, and understood the Pre-reads of this course
You have 3 weekends of time that you can dedicate towards upskilling
You have a Gemini or OpenAI or Anthropic API access
3 weeks • 12th September, 2026
9:00 am to 12:00 pm IST on Saturdays and Sundays
During the live sessions, you will
Enroll Now
Note: This is a discounted, non-refundable price. If you prefer the refundable pricing option, you can select it while making the payment.
YOU'LL GET
Live Classes on Weekends IST
Lifetime access to the cohort recordings
Lifetime access to the Network and Community
Open forums and interaction with the cohort
Doubt resolution during and post live sessions
2 days no-questions-asked refund policy
Language of communication will be strictly english
You will be provided with an invoice with GST details and a certificate of completion, which you can use to get the course fees reimbursed by your employer. If you have questions or need any clarifications before enrolling, please
If you choose recordings, you will
Enroll Now
YOU'LL GET
22+ hours of recordings of my best cohort
Lifetime access to the cohort recordings
Lifetime access to the Network and Community
Doubt resolution on Discord
2 days no-questions-asked refund policy
Language of communication will be strictly english
You will be provided with an invoice with GST details and a certificate of completion, which you can use to get the course fees reimbursed by your employer. If you have questions or need any clarifications before enrolling, please
The primary objective of this program is to make you comfortable at building systems that are scalable, fault-tolerant, and reliable. But here is what you could reap out of it.
Go beyond concepts and learn how to design, implement, and ship applied AI systems that work in real production environments.
Learn the tooling, patterns, and engineering tradeoffs required to build robust AI applications.
Develop the ability to design complete AI systems including data flow, evaluation, reliability, and failure handling.
Learn proven design patterns and architectures that you can directly apply to your own projects at work.
Every concept is backed by code and prototypes so you can see how things work in practice.
Build the skills needed to work effectively with rapidly evolving AI tools and frameworks in industry.
Position yourself to take on high-impact AI projects and stand out as someone who can deliver, not just experiment.
Join a group of serious engineers building AI systems and learn from shared experiences and discussions.
People from all over the world have taken my courses.
Folks belonging to some of the best companies and high thriving startups have taken this and other courses, the list includes the likes of
Here are some of the videos that will give you a peek into my teaching style how I teach and the depth I go into
I am a software engineer and engineering leader passionate about applied AI and databases. Currently, I am a Principal Engineer II at Razorpay, working at the intersection of Data and AI, where I am building Agent Studio, which enables merchants to create AI agents while we provide the infrastructure and harness to run them reliably. I am also working on the Customer Data Platform and leading Merchant Developer Experience.
I am also building px0.ai, an open-source utility to build custom agents and workflows using natural language, then run them on a schedule.
Previously, I was a Staff Engineer at Google, where I worked on GCP Memorystore and GCP Dataproc. On the side, I am building - DiceDB, a fork of Valkey with multi-tiering and query subscriptions. I also spend time on independent research, publishing my work on ArXiv. My areas of interest include databases, approximate algorithms, and distributed systems.
I build a lot of side projects — some take off, some don't, but building them is always fun.
In 2024, I took a leap of faith and co-founded Profile.fyi, which was later acquired by Mercor. I was part of Amazon's Fast Data Team, where I worked on cold tiering of hot data and developed a seamless query interface across all storage tiers.
I held engineering leadership positions (both IC and management) at Unacademy, where I built, grew, and led Search, Site Reliability Engineering (SRE) teams, and Data Engineering teams. I hold a total of 12+ years of experience in scaling backend services, taking products and teams from 0 to 1, and beyond.
I keep diving deep into engineering details and share my learnings through 166+ blogs, socials and, 250+ no-fluff engineering videos on YouTube, breaking down System Design, Database Internals, Microservices, BitTorrent Internals, Redis Internals, Research Papers, and the trade-offs behind real-world production systems.
From the reviews and feedbacks I gathered, here are a few key things that worked for folks who took my courses.
mental models and frameworks
structured and well organised
quality and non-repetitive content
minute implementation-details
open ended discussion
much more than blogs we find
I have known Arpit for the last 5–6 years, and he has been a friend, mentor, and someone I have learned a lot from over the years.
I really enjoyed this course because it explains AI and LLM concepts in a very practical and easy-to-understand way. What I liked most is that it is not just about how to use LLMs, but about how to build reliable AI systems that can actually work in production.
He covered everything from:
prompting, RAG, tool use, and agents
multi-agent systems, memory, evaluations, safety, observability, caching, and production challenges
The system design discussions and hands-on prototypes made the concepts much easier to understand.
I especially liked how Arpit connects the theory with real-world engineering problems and constantly makes you think about things like reliability, scalability, tokens, context, memory, cost, and what can go wrong in production.
Having known Arpit for years, I can say that this course reflects what I have always liked about his way of teaching, he makes complex things simple without losing the technical depth.
If you want to learn how to build reliable AI systems and understand what it takes to make them work in production, I would definitely recommend this course.
Arpit's Applied AI course was one of the most practical and engaging learning experiences I've attended. The sessions were highly interactive, with consistently high engagement that kept everyone actively involved rather than passively listening.
What stood out the most was how seamlessly the course blended core system design principles with applied AI concepts. Instead of focusing on niche AI use cases, the system design patterns covered were extremely generic, robust, and applicable across almost every business domain, making the learnings immediately actionable.
Another highlight was the exposure to cutting-edge AI tools and workflows that are genuinely useful in day-to-day work. The course struck the perfect balance between foundational concepts and hands-on application, ensuring that participants walked away with both a strong mental model and practical skills they could start using right away.
Overall, this was an exceptionally well-designed course that made applied AI feel accessible, practical, and relevant. I would highly recommend it to anyone looking to build real-world AI systems and understand how to integrate AI effectively into modern software architectures.
The Applied AI Masterclass by Arpit has been a great learning experience. What really stands out is how he explains not just the concepts, but how things actually work in production...what issues and challenges come up, and how to solve them with the right system design.
His prototypes were especially helpful in understanding what happens behind the scenes in an AI system and how different components work together. It also gave me a much better understanding of the challenges involved in building AI systems at enterprise scale and how to approach them.
Overall, a very practical and insightful masterclass for anyone who wants to go beyond AI concepts and understand how to build reliable production AI systems.
Coming from a data background and having recently completed my graduation, I was out of the industry for almost two years.
During that time, the whole Agentic AI and Applied AI space grew really fast, and it became difficult for me to gather and connect all the scattered information.
Arpit's course gave me exactly what I was looking for, but in a much more structured way.
What made the learning even more valuable was his industry experience, because the concepts were not just theoretical they were explained with a practical industry perspective.
That is what made this course and the overall learning experience special for me.
Enrolling in Arpit's cohort was a standout learning experience.
Coming from a Principal Database Engineering and platform background, I had explored the LLM space but my understanding was fragmented. I knew the concepts, but lacked a clear architectural mental model for how they come together in production.
This cohort changed that. Arpit goes far beyond APIs and tools, diving into:
RAG
Multi-agent systems
Workflow engines
Agentic applications
...while keeping the focus firmly on real-world engineering—cost, scalability, governance, security, guardrails, and trade-offs.
What I valued most was his ability to cut through the AI hype and approach everything from a practical engineering perspective. I walked away with a much clearer mental blueprint for building production-grade AI systems.
Thank you, Arpit, for the passion, depth, and energy you brought to the cohort!
I recently completed the Applied AI course by Arpit and it was a great learning experience.
Arpit has actually built the kinds of systems he teaches. Every concept is grounded in real production experience, with practical insights into what you can expect when deploying agentic systems at scale.
The course starts with the fundamentals of agentic systems and gradually dives deep into topics such as:
Tool usage and tool calls with LLMs
The factors that influence agent outputs and techniques to improve output quality
Memory architectures, memory management, and memory write strategies
Evals—what makes a good Evals framework and practical strategies for building them, which definitely helps build confidence in developing agentic systems.
In every session Arpit made sure to involve participants in discussion and constantly challenged us to think deeply about system design, production behaviour, failure modes, and the guardrails required to build reliable systems. More than specific techniques, you will learn strong mental models for designing and reasoning about these systems. Absolutely a fun experience( yes we did crack some nerdy jokes)
This was my first cohort with Arpit, and I must say he has an exceptional teaching style. He is one of the most authentic, thoughtful and genuinely curious people you'll interact with.
With so much noise around AI today, Arpit's cohort is a high signal, no-fluff learning experience. If you're serious about understanding how to build reliable agentic systems and want clarity how to go about building them, I highly recommend taking the course.
AI is evolving at an incredible pace, so I wouldn't wait too long to take this course.
Happy learning! And trust me, Arpit's cohort really is a happy learning experience.
P.S. - Don't get scared by the pre-reads. My recommendation is to work through them one by one as soon as you get access to the course.
If you've been to a System Design cohort by Arpit, you know exactly what to expect i.e great discussions and loads of learning.
If you haven't, let me guide you through my experience of the Applied AI Cohort.
Its 3 weeks of super fun discussions, which dive deep into designing Agentic Systems, starting from the basics of what an agentic loop contains, to diving into what it takes to run reliable Agentic systems at scale.
Arpit made you think deeply about how your agentic systems can behave in production, reason about WHY that might happen, and HOW you could design your agent loop and guardrails to prevent that. He dived into Evals, different strategies that can be used while creating Evals which allows you to gain more confidence in the Agentic Systems you're building.
As with any cohort experience with Arpit, we had folks sharing their insights of:
how they're building Agentic Platforms
tools they're using
issues they are facing
which makes the sessions even more informative!
If you're trying to build Agentic Systems and want a solid mental model of:
how they work
how to design them
how to overcome the real-world operational challenges of running them in production
this cohort is absolutely for you! Highly recommend the cohort!
I would like to sincerely thank Arpit for putting together such a well-structured and insightful Applied AI Masterclass.
Before attending the course, I had many questions about how AI solutions are actually built in the real world. While there is plenty of information available online, it was difficult to connect the dots between concepts and practical implementation. This masterclass provided exactly that missing perspective. It helped me understand not only what can be built with AI, but more importantly how to think about designing AI-powered systems from identifying the right use case to choosing the appropriate architecture and incorporating solid system design principles.
One aspect I particularly appreciated was that every session had its own unique value and practical takeaway. The content was thoughtfully organized, balancing fundamentals with real-world applications, making even complex topics easy to understand. The emphasis on engineering thinking, trade-offs, and production-ready AI systems was especially valuable.
Overall, it was a fantastic learning experience, and I truly appreciate the time, effort, and knowledge that Arpit shared throughout the course.
Thank you once again for an exceptional learning experience.
Arpit is a rockstar! Attending the Applied AI course was genuinely fun. The best part is the flow of the sessions: you start with concepts and prototypes, then move into brainstorming and designing for a real-world problem. The problems are well-framed, and Arpit does a great job of connecting the dots so that every concept is covered along the way.
It was incredibly helpful to see how AI systems like coding agents and RAG actually work behind the scenes, and how to reliably build, deploy, test, and observe them in production.
The Q&A at the end of each session is worth waiting for. Just sticking around and listening to other people's questions taught me a ton. Some of the discussions were genuine eye-openers.
Huge thanks to Arpit for putting together such a high-value course :)
Arpit, thank you for being such an inspiration throughout the course.
Every Friday I'd look forward to the weekend because I knew there would be another session that would completely change how I thought about Agentic AI. More than just the concepts, it was the way you:
broke down real-world systems
shared practical insights
actually built prototypes instead of staying at the theory level
Every discussion sparked new curiosity and made me want to go build something on my own. The course didn't just teach me about agents—it changed the way I think about designing AI systems.
Thank you for keeping everything so practical and for sharing your experience so openly. I learned a lot, and I'm genuinely excited to keep building in this space.
The Applied AI cohort is the counterbalance to the AI psychosis out there as it pulls you out of vibe-coding territory and puts Arpit Bhayani in your corner to help you navigate the biggest shift in software since cloud compute and big data.
True to form, he teaches everything from first principles rather than treating AI as some magic wand. You learn to apply real software design fundamentals to "tame the lion" (as he'd say) and get results that actually holds up for your use cases.
I'd strongly recommend this cohort to anyone with the judgment and taste to use AI for leverage and not be the ones who let it replace them.
Two things I like best about Arpit's courses:
First, The brainstorming sessions which puts us on our feet, makes us think through the systems not just from an engineering lens but also from a product & business PoV
Second, The QnAs where every single question is answered logically & practically.
Thank you so much, Arpit for kindling my curiosity yet again with your amazing course.
I have just started working in the Generative AI space and was looking for a course that could give me in-depth, hands-on knowledge to build good AI product features. There are multiple things I really liked about this course:
The way the course is segmented gives you a very clear understanding of each topic. Every concept is explained with tons of practical examples, and the way Arpit narrates them makes it easy to follow.
The best part of the course is the questions raised by the attendees. Many experienced engineers actively participate in the discussions, share their real-world experiences, and ask practical questions. Those discussions make the sessions much more engaging and help us understand the concepts from different perspectives.
Arpit's way of demonstrating problems is unique. He shows how we can validate and experiment with concepts using simple setups instead of requiring a complex environment. He constantly encourages everyone to do hands-on practice, which I believe is the best way to truly learn Generative AI.
Arpit was phenomenal in explaining real problems with ease & his teaching style will never bore you.
I'm already working in AI domain but as all know, learning never stops. I was very eager to see how others are using/incorporating AI in other industries/verticals.
I would recommended to all the AI folks out to try it out at least to get upto speed with market happenings. All the best!
One of the best LIVE cohorts I've attended so far.
Having been part of both Arpit's System Design and Applied AI cohorts, his approach to problem-solving and real-world LIVE interactions are truly unmatched.
Each discussion with Arpit pushes me to look at software problems and solutions at a whole new level and leaves me more convinced that the software industry holds more promise and opportunity than ever before <3
I just finished the Applied AI Masterclass by Arpit Bhayani, and the one-line review is this: it treats AI agents as distributed systems, not as a prompting exercise.
Anthropic now calls the move past prompt-crafting "context engineering." This course goes a layer further.
Some things we covered:
Checkpoint and resume for agent loops
Adversarial evals and LLM-as-judge
RAG over 10M docs with hybrid search, re-ranking, and query rewriting
Self-evolving agents
26 prototypes, 7 system designs, all in Python.
I have been watching Arpit's videos for many years, and he is an excellent teacher. He always goes deep into technical topics, drawing on his own experience and first-principles thinking.
I would recommend this course to anyone looking to get a grip on Applied AI.
I was able to use this as a one stop solution to relieve myself from all the Applied AI FOMO I had and this course really helped me ground myself to the State Of The Art without diving into too much jargons.
The best part is that Arpit has practical experience of building the things that he was teaching and the discussions weren't just situational (with no overlap with reality).
The brainstormings were very involved and Arpit has his way of pushing through the thinking process.
Hoping to attend another much more advanced cohort in the future! Thanks, more power to you :)
It was really amazing and interesting from starting of the course, learned alot of ways to tackle a specific problem. All the prototypes and implementations are based on practical thinking, covered most of the edge cases to make it production ready. The way of solving a particular problem is simple and very practical.
Thank you so much for bringing this cohort. As always you are a great mentor that inspires to think highly practical and build prototypes around the system to understand nuances and trade-offs. I become more curious after completing this cohort, Thanks again.
Arpit is incredibly clear.
I used to jump straight into mapping out backend architectures. Now I actually hit pause. His sessions rewired how I approach high-level design.
Before drawing a single moving arrow, I sit with the product requirements to nail down the real constraints.
I'd honestly rewatch every single video.
Arpit is among the best in the industry. I have been following him for a while, and enrolling in this Applied AI course was a great experience for learning how to build production grade agentic systems.
I will definitely recommend that people enroll.
Thanks, Arpit, for keeping the engineering and agentic curiosity alive. 10 out of 10 from my side!
And on a side note, keep doing what you are doing you make a difference to many.
Arpit explains everything in a very clear and practical way. He uses great prototypes and real-world examples, which makes even complex AI concepts easy to understand.
I especially liked his approach to designing good systems with practical and better solutions. He also takes the time to solve doubts and explain concepts clearly whenever needed.
Overall, a very valuable and enjoyable learning experience.
The course is informative in terms of the concepts that runs behind the AI and the things we use everyday.
It might be good to have a prior knowledge of system design concepts and implementation is everything. The course is only worth it if we try the exercies and implement.
Overall great course and builds curiousity throughout!!
You can always drop me an email at arpit.masterclass@gmail.com for other questions.
I will be teaching the entire course in english and all the brainstorming with participants will be conducted in english.
This course is for any engineer who wants to learn the topics in the curriculum. Please go through the curriculum and decide if it covers what you want to learn.
Yes. I will be teaching the entire course online and live over Zoom.
The Live Classes will happen on Saturdays and Sundays as per the time mentioned on the webite with a possible extension of 30/45 mins.
Talk to your manager and check if they can sponsor this course. The invoice that will be issued is a legally valid and sound invoice that can be used for any kind of reimbursement. Note: I will just share the invoice and certificate with you; and I will not be involved in any kind of reimbursement process, communication, or follow-ups from your finance and legal teams, it is between you and your employer.
You can download the invoice from the course portal. No invoice will be issued if you claim your refund. If you claim a refund, no invoice will be issued.
If you want invoice in the name of your employer or a business for them to claim GST Input Credit (not TDS), drop me an email at arpit.masterclass@gmail.com. Note: I will not be involved in any kind of reimbursement process, communication, or follow-ups from your finance and legal teams, it is between you and your employer. I will not be dealing anything w.r.t TDS.
No. There will not be any 1:1 mentorship sessions for this course.
Yes. I would recommend you implement the core of every single system we discuss ensuring you apply what you learn.
No. But it is advisable that you complete them to get a better understanding of the system, algorithm, and business logic.
Due to time constraints, it is not possible to implement every system; it is recommended that you self-implement the system and understand the low-level details. The course will definitely cover systems from every aspect.
Every single Live Class will be recorded, and you will be given lifetime access to it.
No. You will get lifetime access to the cohort you are part of and its recordings, or the recordings that you purchased.
All Live Classes will be online, over Zoom, and all you need is an internet connection to attend the live sessions.
The entire course including Live Classes will be conducted by me, Arpit Bhayani, no external TAs, mentors, etc. You will get to learn everything from the horse's mouth.
The course prerequisites are mentioned on this page.
Yes. An invoice will be issued to you with all the legal and necessary details. This means your employer can choose to process this invoice and provide reimbursement.
Yes. You can download it from the portal. The moment you download the invoice from the portal, you become ineligible to claim a refund, even if your refund window is not over.
Yes, we support Credit Card, Debit Card, UPI, and Credit Card based EMIs having a duration of 3 months, 6 months, 12 months, and 24 months as offered by Razorpay.
I track the browsers and devices from which the course is being accessed and if I fnd anything suspicious, I hold the complete right to revoke the access of the course and not offer any refund.