Sikkanam - AI-Powered Tamil Nadu Travel Planner & Budget Guide

Sikkanam is an AI-powered travel planner built for Tamil Nadu. It helps travelers discover destinations, generate day-wise itineraries, estimate trip budgets, explore hotels, and plan routes with greater transparency. Unlike generic travel tools, Sikkanam combines destination intelligence, budget planning, route guidance, and an AI travel assistant in one platform. Whether you're planning a weekend getaway or a multi-city trip, Sikkanam helps you travel smarter, spend better, and explore more.

Add a comment

Replies

Best
Hey hunters! 👋 Thank you so much for checking out the launch of Sikkanam! 🎒🗺️ 💡 The Inspiration The inspiration for Sikkanam came from a common frustration: trying to use popular AI tools to plan a budget trip. The AI would construct beautiful stories, but when it came to numbers, it completely fell apart—recommending expensive 1st AC sleeper train classes for a tight ₹5,000 budget, hallucinating non-existent bus routes, or claiming a 500 km journey only takes 3 hours. Budget travel requires strict, realistic parameters. If a travel budget is based on guessed numbers, the trip falls apart. ❌ The Problem We're Solving AI hallucination in travel logistics. We wanted to build a planner that guarantees 100% cost transparency and geographic accuracy. We wanted a tool that actually knows that: Trains from Chennai to Salem or Coimbatore take 5–8 hours (not 3). The Nilgiri Toy Train runs once a day at 7:10 AM, and TNSTC buses are the primary budget-friendly backup. Regional food recommendations should be authentic and verified rather than guessed. ⚙️ How Our Approach Evolved We initially started with a standard LLM chatbot, but quickly realized that generative AI is inherently bad at strict math and real-time transit timetables. To solve this, we pivoted to a hybrid architecture: The Deterministic Cost Engine: We built a local rule engine that handles all calculations (transit fares, room occupancy splits, entry fees) mathematically. The AI is strictly barred from estimating costs. Official Brochure RAG: We parsed the official Tamil Nadu Tourism Department brochures (Heritage, Beach, Ecotourism, and Culinary guides) into a structured JSON database. The AI Companion uses this to retrieve authentic spots (like Tada Falls hikes or Manapparai Murukku origins) and feeds it directly to the user. Dual API Fallback & Cache: We wrapped our API in a dual-LLM fallback (Groq Llama 3.1 + Gemini 1.5 Flash) and pre-seeded local caches to make home-page queries load instantly (0ms latency) and survive rate limits. We'd love to hear your feedback on the planning engine! Let us know what destination you'd plan next! 🚀