Projects using Material UI in PunjabProjects using Material UI in Punjab
Projects
People
Cover image for Keyword Finder, AI SEO Keyword
Keyword Finder, AI SEO Keyword Research & SERP Intelligence SaaS Built a distributed SEO research platform that expands a single seed keyword into thousands of opportunities combining Google Ads volume data, live SERP analysis, NLP competitor scoring, domain authority metrics, and AI content brief generation into one unified workflow with no CSV exports or tool-switching. SEO agencies were stitching together four or five separate tools a keyword research tool, a SERP analyser, a domain authority checker, an NLP scorer, and a rank tracker and manually reconciling their exports in spreadsheets. Scaling that workflow across multiple clients and thousands of keywords every month was unsustainable. Keyword Finder is built on Django 5, FastAPI, Flask microservices, React 18, and GPT-4o. The platform supports 4 keyword generation modes — Seed, URL, Competitor, and Bulk. In Seed mode it builds prefix, suffix, and alpha combination lists, fires them to Google Autocomplete in parallel with rotating proxies, then runs a second deep expansion round before fetching volume, CPC, and competition data from the Google Ads API via OAuth2. SERP analysis jobs are distributed across a dedicated Flask scraping microservice fleet with round-robin load balancing and a four-tier fallback chain direct, Bright Data, scraping.narf.ai (http://scraping.narf.ai), and CrawlBase — ensuring near-100% page retrieval success regardless of target site protection. Each competitor URL is processed through a full NLP pipeline using YAKE and NLTK, with domain authority pulled from Moz and Majestic via DomDetailer and cached locally to avoid redundant API calls. The Ranking Recipe lets agencies define custom Easy, Medium, Hard, and Easiest thresholds across 12 metrics turning keyword prioritisation from guesswork into a data-driven decision. AI content briefs are generated using GPT-4o and Cohere with full keyword context, cutting brief writing time from hours to minutes. Real-time job progress is delivered via Django Channels WebSockets, with Celery and Redis managing the async scraping queue. Supports 200+ Google country domains for multi-market research. Tech Stack: Django · FastAPI · Flask · React · GPT-4o · LangChain · Celery · Redis · PostgreSQL · Google Ads API · Bright Data · YAKE · NLTK · Moz · Majestic · Material UI · Redux
1
86