Kishan Thanki

AI-Powered Resume Analysis Platform

An intelligent platform that evaluates candidate resumes against job descriptions, ensuring transparent scoring by citing exact quotes from the candidate's resume.

• Provides instant candidate evaluations by extracting key job requirements and cross-referencing them with resume claims.
• Eliminates AI hallucinations by forcing the LLM to cite exact quotes from the candidate's resume for every score it generates.
• Built for scale using a dual-language architecture: A secure Go API gateway that delegates heavy AI workloads to a high-speed Python gRPC Engine.
• Enforces strict JSON schema boundaries via Pydantic, ensuring downstream HR systems receive perfectly structured data every single time.

Static Analysis CLI Tool

A static analysis utility for mapping complex Python project architectures, leveraging AST to detect structural loops and generate visual dependency graphs.

• Architected a Python CLI tool to analyze codebase architecture, mapping complex dependency structures and exposing circular architectural coupling.
• Engineered a static analysis engine utilizing Python’s ast module to accurately resolve absolute, relative, and dotted imports without executing arbitrary code.
• Implemented a depth-first search (DFS) algorithm to detect dependency cycles, generating isolated subgraphs and visual dependency maps via Graphviz (PNG/SVG) and HTML.

Scalable Campus Recruitment Backend

A centralized campus recruitment system utilizing Django REST Framework to automate placement workflows, manage production configurations, and securely isolate user data.

• Architected a scalable backend for a centralized campus recruitment system using Django REST Framework, automating placement workflows for administrators, companies, and students.
• Implemented secure, stateless authentication using HTTP-only cookies and JWTs to mitigate XSS vulnerabilities, coupled with a robust Role-Based Access Control (RBAC) and object-level permissions model.
• Engineered a globally consistent API architecture by developing custom middleware and BaseViewSets for unified JSON responses and centralized exception handling.
• Deployed the production-ready backend on Render, managing environment configurations, securing database connectivity, and integrating Cloudinary for optimized media storage.