Why the Tech Stack Conversation Is Overrated
Ask ten developers what tech stack a startup should use and you'll get twelve opinions. The reality is that the tech stack matters a lot less than whether the team executing on it is any good. Twitter ran on Ruby on Rails for years at massive scale. Shopify still does. WhatsApp served 450 million users with a team of 35 engineers. The tools weren't the limiting factor — the engineering quality was.
That said, a genuinely bad choice — picking a niche language with a tiny talent pool, or an over-engineered architecture for a product that has zero users yet — can create real problems. Here's our actual recommendation for Indian startups in 2026, based on what we've seen work and fail.
For Web Applications: Next.js + Node.js + PostgreSQL or MongoDB
This is our default recommendation for most web products — a B2B SaaS, a marketplace, an internal tool, a customer portal. Here's why:
- Next.js (React framework): Handles routing, server-side rendering, API routes, and image optimisation in one package. The best default for new web apps. Large developer community in India means easier hiring.
- Node.js: Same language (JavaScript/TypeScript) on frontend and backend. Fast to build with, huge ecosystem, widely known in the Indian developer market.
- PostgreSQL for structured data with complex relationships (financial data, inventory, multi-tenant SaaS). MongoDB for flexible document-shaped data (content, user profiles, catalogue). Don't mix them unless you have a clear reason.
- Vercel or AWS: Vercel for fast deployment and zero infrastructure management in early stages. AWS when you need more control or cost optimisation at scale.
For Mobile Apps: Flutter (or React Native)
If you need both Android and iOS, Flutter is our current recommendation. One codebase, near-native performance, strong Google backing, and the ecosystem has matured significantly in the last two years. The talent pool in India is growing quickly — Flutter developers are easier to find in 2026 than they were in 2022.
React Native is a valid alternative, especially if your team already knows React and wants to share code between web and mobile. It's more mature, has a larger community, but has more platform-specific edge cases than Flutter.
Native (Swift for iOS, Kotlin for Android) only if your app has deep platform-specific requirements — camera/AR features, complex system integrations, or performance requirements that cross-platform frameworks genuinely can't meet.
For AI-Powered Features: Python Where It Makes Sense
If your product involves machine learning, data processing, or AI model integration, Python is the right tool for those specific parts. The ML ecosystem — PyTorch, scikit-learn, Hugging Face, LangChain — is Python-first. This doesn't mean your whole stack needs to be Python; it means having a Python service or microservice for AI-specific work, with the rest of your stack in JavaScript/TypeScript.
For calling existing AI APIs (OpenAI, Anthropic, Google Gemini), any language works fine — you're making HTTP calls, not writing ML code. No need for Python just because you're using an AI API.
What to Avoid as a Startup
- Microservices from day one: Unless you have 20+ engineers and very clear scaling requirements, a monolith is the right starting point. Microservices add operational complexity that slows small teams down significantly.
- Niche languages for the core product: Elixir, Haskell, Rust, Go can be excellent choices — but if the team doesn't already know them deeply, the learning curve and small Indian talent pool will slow you down. Use them when the specific advantages (concurrency, systems programming) are genuinely needed.
- Over-engineering infrastructure early: Kubernetes, service meshes, and complex CI/CD pipelines are solutions to problems you don't have yet. Start simple: a VPS or PaaS, a single database, a basic deployment pipeline. Add infrastructure complexity when the pain of not having it becomes real.
The Most Important Variable: Team Quality
A strong team using a decent stack will outperform a weak team using a "perfect" stack every time. When evaluating tech decisions, ask: does our team know this well? Can we hire for this in India? Will this still be a good choice in three years? The answers to these questions matter more than benchmark comparisons.
At Dharmsy, we've settled on a stack we know deeply — Next.js, Node.js, MongoDB/PostgreSQL, Flutter — and we build well with it. If you're a startup figuring out what to build and how to build it, we're happy to have that conversation.