Applied AI Engineering is a red hot subject in Software engineering and in this video and in this video series, I will be explaining, how to get into this field to create Applied AI products and provide value to our users with the technical concepts we have learnt till now.

In this video I have mentioned how you can implement the backend for you Applied AI Engineering product in a brief way (as explaining every feature and tools I have used and might use will take multiple hours and even this video is 50+ minutes long lol).

I have explained how you can implement llm inference, data engineering to stream data into your database etc.

00:00 – Intro: Backend for AI Engineering Concepts
00:41 – What This Video Covers: GPU Inference, VLLM, MLOps, Embeddings
01:20 – Second Brain Labs Overview (Not Sponsored)
02:31 – Manual Sales vs AI Agent Automation
04:25 – WhatsApp for India: Why It’s the Key Channel
05:33 – How SLB Automates Campaigns with AI
07:24 – Coupon Tracking with Stripe and Lead Attribution
09:18 – Behind the Scenes: How They Might Have Built It
10:25 – Hosting Open-Source LLMs with VLLM + Model.com
12:32 – Deploying AI APIs Like Traditional Web Servers
15:18 – Why Self-Hosting Beats OpenAI for Enterprises
17:21 – GPU Ownership vs Cloud Inference (3 Options)
20:04 – Real Enterprise Pitch: Private LLMs for Security
22:29 – Ingesting Data into AI Agents: Knowledge Base + Leads
25:03 – Data Engineering with ETL and Streaming
28:10 – Scalable Fan-Out Architecture for Real-Time AI
29:53 – Netflix Case Study: ML + RAG Pipelines
31:14 – Embeddings + Vector DBs: Making Agents Smart
34:35 – Querying Lead Data via Vector Search
36:31 – Business Logic: Timing Personalized Outreach
39:02 – Conversions Powered by AI Agents (SLB Example)
41:21 – Agents Work 24×7 Across Timezones
44:23 – Sales Cost vs AI Agent ROI
46:55 – Upload, Connect, Automate: Product Simplicity
48:52 – Embedding Dimensions & Vector Tradeoffs
50:00 – Trufy Case Study: Fraud Detection with AI Agents
53:34 – Governments Use Agents Too: Time + Value
54:42 – Final Thoughts: Real Companies, Real Use Cases

—–

Links mentioned in the video:

Terrific resource to start – https://janvikalra.substack.com/p/how-to-evaluate-model-infrastructure

Understanding AI – https://leerob.com/ai

Agno AI Agents – https://docs.agno.com/introduction

Second Brain Labs – https://secondbrainlabs.com/
(The product which we are going through in the video to understand how they MIGHT have implemented the Automated Sales AI Agent to provide value to their customers)

Hugging face: https://huggingface.co/docs/transformers/en/model_doc/llama

Inference Providers

Modal: https://modal.com/

Cerebrium: https://www.cerebrium.ai/

Together AI: https://www.together.ai/

vLLM Inference to host the open source AI modals: https://docs.vllm.ai/en/latest/

AJVC Funded Startups: https://www.ajuniorvc.com/portfolio-companies/trufides
(I have explained how they MIGHT have used the concepts mentioned in the video to created AI Agents to provide value to their customers)

—–

You can email me at for further queries: gauthamvijay495@gmail.com

You can find the source codes for my videos in topmate: https://topmate.io/gautham

If you have any other questions, please leave it in the comments or contact me via Twitter (X) over DM’s.

Here is my Twitter (X) profile: https://x.com/gautham_vijay_

source