In 2026, Programming Evolves: From Typing to Thinking
System-level engineer building reliable backend systems with a focus on performance, correctness, and real-world constraints. I work across APIs, databases, networking, and infrastructure, enjoy understanding how systems behave under load and failure, and write to break down complex backend and distributed-systems concepts through practical, real-world learnings.
How I Built BanglaCode — a Banglish Programming Language — in a Sunday Vibe 🚀
Sunday, 8 Feb 2026 — the world already feels different from what it used to be even 3 years ago. AI in 2026 isn’t just an autocomplete or helper; it is a reasoning partner. Today I experienced that shift firsthand.
Before this morning, I asked myself a simple question: Can code speak Bangla? Not just comments or variable names, but real logic written in Banglish — a syntax that feels like my mother tongue, not English. So I started building BanglaCode — a Banglish-styled programming language powered by GoLang 🐹.
For context, the language I envisioned was not just a gimmick — it has structured syntax, keywords, and rules. It supports typical programming constructs like variable definitions, functions, loops, and expressions, but written in Bangla transliteration — making programming logic feel natural for Bengali-speaking thinkers. I drafted grammar rules inspired by syntaxes you see in languages like C/Go, but adapted to Banglish expressions (taking cues from the SYNTAX document I created in the GitHub repo). Although the syntax file itself is long and detailed, the core idea was: make programming feel conversational rather than foreign.
💡 An example from the BanglaCode syntax mindset:
Instead of writing if (x > 5) { … }, you might express it in a Banglish style that speaks closer to your intuition. This is not just code — it’s language that thinks like you do.
I didn’t just scribble pseudocode. I went into Go, built the lexer, parser, and interpreter logic — I vibe-coded it. Because in 2026, the AI tools I use understand patterns, logic, and intent — not just text matching. They help me reason with the language, not just autocomplete lines.
By afternoon, I had a working prototype:
✔️ BanglaCode compiles
✔️ Test code runs
✔️ Syntax rules from my design are respected
✔️ It even handles basic control flows and variables
…and most importantly, it felt alive — like it could grow into something real, not just a weekend project.
What did this teach me about programming in 2026?
👉 AI doesn’t replace humans — it amplifies clarity.
If your logic is weak, AI amplifies confusion.
If your logic is strong, AI becomes unbelievably powerful.
Today I didn’t just write a language; I experienced the future of programming — where AI helps you translate an idea directly into structured logic without getting bogged down by foreign syntax. In 2026, programming isn’t about memorizing keywords — it’s about formulating thought clearly enough for a machine to understand and execute.
And on a Sunday? That’s a vibe. 😄🔥