Intellicruit
Intellicruit: AI that reads resumes, scores candidates, and schedules interviews—so you hire the best, faster, while empowering HR teams and helping candidates shine.
Created on 26th May 2025
•
Intellicruit
Intellicruit: AI that reads resumes, scores candidates, and schedules interviews—so you hire the best, faster, while empowering HR teams and helping candidates shine.
The problem Intellicruit solves
Recruitment is often time-consuming and manual, requiring HR teams to review hundreds of resumes and schedule interviews across different time zones.
Bias in candidate evaluation can lead to unfair hiring decisions and lack of diversity in the workplace.
Candidates struggle with inconsistent feedback, limited interview preparation resources, and unclear expectations for roles.
Existing systems are often fragmented, requiring multiple tools for resume parsing, scoring, scheduling, and feedback, leading to inefficiency.
Intellicruit solves these problems by:
Automatically parsing resumes (PDF, DOCX, images) to extract structured data like skills, experience, and certifications.
Scoring candidates objectively based on alignment with job descriptions, highlighting strengths and areas for improvement.
Providing instant feedback to candidates through AI-driven mock interviews, helping them prepare better.
Automating interview scheduling by matching HR and candidate availability, reducing manual coordination.
Offering AI-powered job recommendations based on resume analysis and market trends, helping candidates find better-fit opportunities.
Supporting continuous learning through course recommendations and certificate verification.
Reducing bias by focusing on skills, experience, and competencies rather than subjective factors.
Intellicruit streamlines the recruitment workflow for HR professionals and provides a fair, efficient, and data-driven experience for candidates.
Ultimately, Intellicruit helps organizations hire the right talent faster and empowers candidates to prepare effectively for technical roles.
Challenges we ran into
Parsing Diverse Resume Formats
Resumes came in various formats—PDF, DOCX, scanned images—which made it difficult to extract consistent data.
Solution: Integrated PyMuPDF and Unstructured (OCR) for parsing, with LangChain to unify context and structure across formats.
Ensuring Fair AI Scoring
Initial AI scoring models showed bias toward certain keywords and roles.
Solution: Refined prompt templates, used MCP (Model Context Protocol), and incorporated feedback loops to improve fairness and accuracy.
Synchronizing Multiple AI Agents
Coordinating Resume Agent, Scoring Agent, Mock Interview Agent, and Scheduler Agent led to data inconsistencies.
Solution: Implemented A2A (Agent-to-Agent) communication protocols and standardized data formats for smooth handoffs.
Mock Interview Timing & Realism
Managing question prompts, recording responses, and analyzing non-verbal cues in real-time was complex.
Solution: Combined pyttsx3 for TTS, OpenCV for video, Whisper for transcription, and custom silence detection for precise flow.
Scheduler Time Zone Handling
Parsing user-provided availability (like "Monday 2 PM") into real calendar dates across time zones was error-prone.
Solution: Used LLaMA3 for NLP parsing and built robust logic to map natural language times to actual time slots.
Managing API Limits & Performance
External APIs like Whisper and Groq (LLaMA3) often hit rate limits during bulk processing.
Solution: Added asynchronous task queues, batching, and retries to optimize API calls.