Screenshots of the finished app will be added here before it goes on sale.
Coming soonAdvanced
AI Agent Application
An agent built with LangGraph that plans multi-step workflows, calls tools and external APIs, and keeps memory between turns. A practical base for building your own agents.
from ₹499 at launch
One-time payment · 3 editions · licence · not on sale yet
For: Aspiring AI engineers
Overview¶
The problem it solves
Agent tutorials stop at one tool call. Real agents need several steps, memory, error handling and limits so they don't loop forever. This app shows all of that in one working codebase.
Who it's for
Aspiring AI engineers · Advanced level
Main benefits
- A working LangGraph agent you can extend with your own tools
- Memory across turns, stored so conversations survive a restart
- Safe defaults - step limits, error recovery, confirmation before actions
- The agent patterns asked about in AI engineering interviews
Features¶
This app is in development. These are the features of the finished app; screenshots and the exact file list will be added before it goes on sale.
What it does
- Chat with an agent that plans and carries out multi-step tasks
- Tools that call external APIs
- Remembers the conversation across turns
- Asks before taking actions that change things
Technical capabilities
- LangGraph state graph with tool-calling nodes
- Tool definitions with typed arguments
- Conversation memory persisted with SQLite checkpoints
- Step limit, retries and error handling for tool failures
- FastAPI endpoint for chatting with the agent
Architecture¶
flowchart LR
U[User] -->|message| API[FastAPI endpoint]
API --> A[LangGraph agent]
A --> L[Groq LLM: plan next step]
L -->|tool call| T[Tools]
T -->|result| A
T --> X[External APIs]
A <--> M[(Memory: SQLite checkpoints)]
L -->|final answer| U
Data flow
- The message reaches the FastAPI endpoint and is added to the conversation state.
- The LLM decides the next step - call a tool, or answer.
- Tool calls run with typed arguments; results go back into the state.
- The loop repeats until the LLM answers or the step limit is reached.
- The state is saved as a checkpoint, so the conversation continues next time.
Backend components
- FastAPI chat endpoint
- Tool implementations and API clients
- SQLite checkpoint store for memory
AI components
- LangGraph state graph (agent and tool nodes)
- Groq-hosted LLM with tool calling
- System prompt with rules for tools and confirmations
Technology stack¶
| Technology | What it does here |
|---|---|
| Python | The language of the whole app |
| LangGraph | Agent graph, state and loops |
| Groq | LLM with tool calling |
| FastAPI | API to chat with the agent |
| SQLite | Stores memory checkpoints |
What's included¶
Everything is delivered as a private download after payment. Nothing paid is ever public on this site.
| What you get | Learner | Career | Developer |
|---|---|---|---|
| Complete source code (ZIP)The whole application: backend, frontend and configuration, organised as a real project. | ✓ | ✓ | ✓ |
| READMEWhat the app does, how the code is organised and how to run it. | ✓ | ✓ | ✓ |
| Requirements filerequirements.txt with pinned versions, so installs are repeatable. | ✓ | ✓ | ✓ |
| Environment example.env.example listing every setting and API key, with safe defaults. | ✓ | ✓ | ✓ |
| Setup tutorialStep-by-step guide from download to a running app on Windows, macOS or Linux. | ✓ | ✓ | ✓ |
| Architecture diagramThe diagram on this page as an image, plus the source to edit it. | – | ✓ | ✓ |
| Architecture walkthroughWhy each component exists and how a request flows through the code. | – | ✓ | ✓ |
| Interview Q&AQuestions interviewers ask about a project like this, with model answers. | – | ✓ | ✓ |
| Resume bullet pointsReady-to-adapt bullet points describing the project for your resume. | – | ✓ | ✓ |
| Docker configurationDockerfile and docker-compose.yml to run it in containers. | – | – | ✓ |
| Test casesAutomated tests (pytest) for the main features. | – | – | ✓ |
| Deployment guideHow to deploy it to a cloud platform, with the settings to change for production. | – | – | ✓ |
System requirements¶
| Python | Python 3.11 or newer |
|---|---|
| Memory (RAM) | 2 GB |
| Disk space | About 500 MB for dependencies |
| Python packages | langgraph langchain-groq fastapi uvicorn httpx |
| API keys and services | A Groq API key (Groq offers a free tier with rate limits) Keys for any extra APIs you connect yourself |
| Operating system | Windows, macOS or Linux |
Choose your edition¶
Learner Edition
₹499 one-time · at launch
Learn how it works and run it yourself.
- Complete source code (ZIP)
- README
- Requirements file
- Environment example
- Setup tutorial
Not available at the moment.
Career Edition
₹799 one-time · at launch
Everything to present it in interviews and on your resume.
- Complete source code (ZIP)
- README
- Requirements file
- Environment example
- Setup tutorial
- Architecture diagram
- Architecture walkthrough
- Interview Q&A
- Resume bullet points
Not available at the moment.
Developer Edition
₹1,499 one-time · at launch
Production extras and a commercial licence for client work.
- Complete source code (ZIP)
- README
- Requirements file
- Environment example
- Setup tutorial
- Architecture diagram
- Architecture walkthrough
- Interview Q&A
- Resume bullet points
- Docker configuration
- Test cases
- Deployment guide
Not available at the moment.
Prices are in Indian rupees, one-time, with no subscription. Not on sale yet: "Notify me" adds you to the launch list for this app.
Licence and support¶
Personal licence
Use the code to learn, run it for yourself and show it in your portfolio and interviews. You may not resell or redistribute the code, or use it in work for clients.
Comes with: Learner Edition, Career Edition
Commercial licence
Everything in the personal licence, plus use in your own commercial products and in work for clients. You may not resell or redistribute the source code itself, as-is or as a template.
Comes with: Developer Edition
- Learner Edition: Help by email if a file is missing or won't open.
- Career Edition: Help by email if a file is missing or won't open.
- Developer Edition: 30 days of email support for setup and code questions, plus help with file problems.
Full wording: Licence information · Refund policy
FAQ¶
When can I buy it?
It's in development and not on sale yet. Leave your email below and you'll get one email when it launches.
Which edition should I choose?
Learner if you want to study the code and run it yourself. Career if you'll present it in interviews or on your resume: it adds the architecture walkthrough, interview Q&A and resume bullet points. Developer if you'll use it in client or commercial work: it adds Docker, tests, a deployment guide, a commercial licence and 30 days of email support.
How will I receive the files?
After payment you'll get a private download of the ZIP for your edition. Paid files are never public on this site.
Does it include API keys or hosting?
No. You'll use your own API keys (see System requirements) and run it on your own machine or server.
Can I use it for client work?
Only with the Developer Edition's commercial licence. See Licence information.
Can I get a refund?
See the Refund & Cancellation Policy. It will be updated with the terms for source code before any app goes on sale.
I have a question that isn't answered here.
Use the contact page or email aiwithpriyabrat@gmail.com. We reply within 2 working days.
Related applications¶
Get AI Agent Application at launch
It's in development. Leave your email and you'll get one message when it goes on sale, from ₹499. No spam.