# QFLOO

> AI engineering courses by Raju Kancharla on LLM evaluation, retrieval, inference and serving, and agents. Hands-on labs, and free lessons in every course.

## Courses

- [Advanced Python](https://qfloo.com/courses/advanced-python): Predict what CPython does with a line of code, choose a concurrency model with measured numbers, and profile and speed up a real program with results you can defend.
- [AI Agents: What They Are and How to Build One](https://qfloo.com/courses/ai-agents-fundamentals): Build a small tool-using agent from scratch, know where agents break, and know when a fixed workflow is the better choice.
- [AI Security and Red-Teaming](https://qfloo.com/courses/ai-security): Threat-model an AI application, break it with prompt injection and tool abuse in a safe lab, apply layered defences, measure attack success with error bars, and respond to an incident.
- [Build Your First AI Application](https://qfloo.com/courses/build-your-first-ai-app): Ship a small AI application that has a test set, error handling and a cost limit, built step by step as a support-email triage tool.
- [Master Competitive Programming](https://qfloo.com/courses/competitive-programming): Solve contest problems with a tested template, a stress-test habit and a toolkit of proven patterns, and run a practice routine built from your own weakness list.
- [Deep Dive into Databases](https://qfloo.com/courses/database-internals): Explain and predict how a database stores, indexes, plans, isolates, logs and replicates data, diagnose slow queries and bad plans, run a safe schema change, and build a small storage engine with a write-ahead log and tested crash recovery.
- [Deep Learning and Transformers from Scratch](https://qfloo.com/courses/deep-learning-transformers): Build, train and evaluate a small decoder-only transformer from raw array code, and explain each part (backpropagation, optimiser, attention, scaling, fine-tuning) with numbers from your own runs.
- [Agent Reliability Engineering](https://qfloo.com/courses/evals-agents): Build an eval harness that tells you, with numbers and error bars, whether a change made your agent better or worse, and blocks the release when it is worse.
- [How AI Is Built: From Data to a Working Model](https://qfloo.com/courses/how-ai-is-built): Follow one model from raw data to deployment, know what each stage costs and where it goes wrong, train, evaluate, release and serve a small model yourself, and see how the same ideas scale to language models.
- [Inference and Serving Engineering](https://qfloo.com/courses/inference-serving): Predict, measure and improve the latency, throughput and cost of serving a language model, and plan capacity for a target, using numbers from your own runs.
- [LLM Performance Engineering](https://qfloo.com/courses/llm-performance): Cut the cost and p95 latency of an LLM feature with measured changes, and prove quality did not drop.
- [Agent Tooling and MCP Engineering](https://qfloo.com/courses/mcp-tool-design): Design, build, test and secure a tool server that an agent can use correctly and that cannot be turned against its users.
- [ML Mathematics for AI Engineers](https://qfloo.com/courses/ml-mathematics): Compute, check and explain the mathematics inside a model (shapes, gradients, likelihoods, entropy and floating point error) well enough to debug a training run from the numbers.
- [Retrieval Engineering](https://qfloo.com/courses/production-rag): Build a retrieval-augmented system whose retrieval and answers are measured, so a change to chunking, ranking or data cannot quietly make it worse.
- [Reading AI Research Papers](https://qfloo.com/courses/reading-ai-papers): Decide whether a paper's evidence supports its claim, rebuild its equations and cost numbers, read the papers behind a technical decision in a defensible order, and reproduce one result with an honest account of the gap.
- [System Design for AI Products](https://qfloo.com/courses/system-design-ai): Design an AI product end to end (requirements, capacity and cost, architecture, agents, retrieval, serving, queues, caching, reliability, tenancy, safety, evaluation and observability) and defend the design in a written document with numbers.
- [What Is AI? A Clear Introduction](https://qfloo.com/courses/what-is-ai): Explain what AI, machine learning and large language models are, what they can and cannot do, decide whether a task suits AI, and run a small honest evaluation of your own.
- [Working with LLMs: Prompts, Context and Limits](https://qfloo.com/courses/working-with-llms): Get dependable results from an LLM, know when it is guessing, and keep a personal test set that catches regressions when you change a prompt or a model.

## AlgoRank

- [AlgoRank](https://qfloo.com/algorank): original practice problems judged in a sandbox for Python 3 and C++, with measured time and memory. 246 problems today; the statements of the free ones are public, the rest are in the paid plan.
- [Cut To Match](https://qfloo.com/algorank/problems/anagram-gap) (free)
- [Template Tag Audit](https://qfloo.com/algorank/problems/balanced-tags) (free)
- [Water Stop Gap](https://qfloo.com/algorank/problems/closest-gap) (free)
- [Audit Window](https://qfloo.com/algorank/problems/ledger-ranges) (free)
- [Two Orphan Parts](https://qfloo.com/algorank/problems/odd-pair) (free)
- [Ticket Range Count](https://qfloo.com/algorank/problems/seat-search) (free)
- [Crane Pairs](https://qfloo.com/algorank/problems/shelf-pairs) (free)
- [Decorated Slots](https://qfloo.com/algorank/problems/tile-floor) (free)

## Founder

Raju Kancharla is the founder of QFLOO and designs its courses.
- He has built many startups and worked across companies of various sizes.
- The courses are designed by him together with expert friends at large technology companies, who review the lessons.
- LinkedIn: https://www.linkedin.com/in/rajukancharla
- GitHub: https://github.com/rajukancharla21
- Contact: raju@qfloo.com

## Pages

- [Courses](https://qfloo.com/courses): All courses
- [Pricing](https://qfloo.com/pricing): Plans and how they work
- [FAQ](https://qfloo.com/#faq): Common questions
- [Privacy](https://qfloo.com/privacy): How QFLOO collects, uses and protects your personal data.
- [Terms](https://qfloo.com/terms): The terms for using QFLOO.
- [Refund policy](https://qfloo.com/refunds): QFLOO sales are final. Read how plans, cancellation, billing mistakes and invoices work.

## Contact

- Raju Kancharla, raju@qfloo.com

## Company

- Legal name: QFLOO AI PRIVATE LIMITED
- CIN: U70200AP2025PTC117413
- Registered address: 7-6-569, Dhanama Nilayam, Rajeev Gandhi Nagar, Guntur, Andhra Pradesh 522002, India

## FAQ

Q: Who is this for?
A: Developers and ML engineers who build on language models and want the depth to measure, design and run them well. The Start here courses assume no AI background.

Q: What does "Learn what engineers at Google, Amazon and OpenAI know" mean?
A: It names the standard we hold the lessons to: depth beyond tutorials, with sources cited, results from code you can run, and evidence before opinion. "L12 knowledge" is shorthand for that depth, not a title or a promise, and levels differ from company to company. It does not say anyone from those companies teaches, reviews or endorses the courses. QFLOO is not affiliated with any of them, and it does not promise a job or a level.

Q: What do I need?
A: Python and git for the hands-on courses. The labs run offline against stand-in models, so no API key is needed.

Q: Where should I start?
A: New to AI: What Is AI?, then Working with LLMs. Shipping LLM features: Agent Reliability Engineering. Curious how models are served on GPUs: Inference and Serving Engineering. Every course has free lessons, so you can look before you choose.

Q: How are the courses made?
A: Sources cited, results from running the lab code, a verified date shown. The author reads, runs and approves each lesson first.

Q: Are all the courses finished?
A: A lesson is published only after the author has read it and run its code. Each card counts the lessons that are open today, and a course that is not open yet says Coming soon.

Q: How do the plans work?
A: Lifetime is one payment with no renewal. Monthly, Quarterly and Annual are subscriptions billed every month, three months or year. The final price, with any taxes, shows before you pay. Checkout opens when payments are switched on.

Q: Can I share my account?
A: No. One account is for one person. Do not share your login, sell access or let others use it. Only one device can be signed in at a time: signing in on a second device ends the first session. This is part of the Terms of Service.

Q: What happens to my data?
A: We keep your email, name and learning progress to run your account, and nothing is sold. The Privacy Policy lists exactly what we store and why.

Q: What if it is not for me?
A: The free lessons need no card, so you can find out before you pay anything.

Q: Can I get a refund?
A: No. All sales are final and there are no refunds, so try the free lessons first. Nothing renews automatically, so there is nothing to cancel. Billing mistakes by us or the payment provider are corrected. See the Refund policy at /refunds.