June 1, 2026, marked my official Day 1 as an AI & Prompt Engineering intern at Virtual Height Pvt Ltd. Under the guidance of our lead mentor Rashmi—an experienced Python developer, Senior AI Trainer, and PhD candidate—the cohort was introduced to the foundational pillars of the Artificial Intelligence landscape and a rigorous cross-platform Python development curriculum.

1. The Theoretical Architecture of AI

We kicked off the technical deep-dive by mapping the core pillars of the field: Artificial Intelligence (the broad science of mimicry), Machine Learning (the statistical logic of training), and Deep Learning (neural networks mimicking biological systems). Rashmi segmented AI classifications based on two vectors: Capabilities (Narrow AI, General AI or AGI, and Super AI or ASI) and Functionalities (Reactive Machines, Limited Memory, Theory of Mind, and Self-Aware Systems).

For our research assignment, we investigated historical milestones toward AGI, such as Hanson Robotics' Sophia. In analyzing Sophia, it became clear that while she represents a sophisticated hardware performance and reactive voice system, she lies far from the soft-logic, open-world reasoning of modern Large Language Models (LLMs)—underscoring how rapidly the field is shifting from rigid robotics toward flexible generative architectures.

2. The Python Compiler Lab & Modern Ecosystem

Following a mid-day break, we transitioned into Python development pipelines. We verified execution mechanics across Windows, macOS, and Linux CLI terminals, comparing command triggers like 'py filename.py' and 'python filename.py'. We cataloged the modern framework landscape: Web architectures (Django, Flask, FastAPI), core Data/ML libraries (NumPy, Pandas, Scikit-learn), and bleeding-edge Generative AI layers (Transformers, LangChain, LLaMA).

3. Programmatic Data Typology

We concluded the technical segment with a hands-on scripting lab covering naming conventions, keyword rules, and datatype primitives. The lab focused on numbers (ints, floats, and complex representations like '3 + 4j'), strings (exploring Python's four literal forms: single, double, triple-single, and triple-double quotes), boolean states, sequence types (Lists, Tuples, Dictionaries, Arrays, Sets), and explicit typecasting mechanics.

4. The Spotlight Breakthrough: A Masterclass Separation

While many in the cohort struggled with basic Python features or naming rules, I completed the datatype casting scripts ahead of schedule. Before closing the session, I presented some of my existing builds to Rashmi. I demoed my custom LLM chatbot—which confidently replies 'Sagar built me' when asked about its origin—followed by my DeepSeek API reverse engineering project. Finally, I walked her through my personal portfolio (sagarithm.in) and digital studio hub (pixartual.studio).

Rashmi was highly impressed, applauding the technical execution while the rest of the room looked on in silence. Having an experienced Senior AI Trainer and PhD candidate appreciate my craftsmanship made my day. It proved that building in public, shipping real-world systems, and moving beyond theoretical learning places you light-years ahead of the traditional curve.

Key Learnings

  • Deep-dive mapping of AI taxonomies from reactive systems to AGI milestones like Sophia.
  • Cross-platform execution mechanics of Python scripting across Windows, macOS, and Linux terminals.
  • Typology, literal structures (4 string forms), and mutability rules of lists, tuples, sets, and dicts.
  • Deploying custom LLM chat interfaces and reverse engineering DeepSeek protocol APIs.

Tools & Stack

  • Python 3.11
  • API Reverse Engineering
  • Custom LLM Chatbot
  • sagarithm.in
  • pixartual.studio

Challenges Overcome

  • Synthesizing high-level theoretical AI capabilities with low-level Python primitive scripts.
  • Maintaining an accelerated building pace within a highly entry-level academic cohort environment.

Task to be Performed

  • Research foundational AI classifications (Narrow, General, Super AI) and functional properties.
  • Perform comparative research on Hanson Robotics' Sophia and modern LLM architecture.
  • Establish local cross-platform Python script run pathways and verify primitive data types.

Related Logs

Day 00

Onboarding & Exploration: Mapping the Core AI & Python Blueprint

Onboarding at Virtual Height, exploring Python-based AI development pipelines, setting up sandboxes, and defining our core AI & Prompt Engineering curriculum.

May 29, 20264 min readRead Log
Day 02

Day 02: Python Control Flow, Functional Lab, and the Birthday Aura Hack

Setting up Python and VS Code envs, implementing conditional systems and match-case control flows, scripting a modular calculator, and presenting a surprise birthday web build to Rashmi.

Jun 02, 20264 min readRead Log