Babasaheb Naik Group name / no.
Branch
College Of Section
year & other details
Engineering 1
HireLens - Gives recruiters a clearer lens.
Problem Statement: Create a system that has pre-
feeded resumes as well as provision to upload
resume. There will be an option to upload job
description as well. Idea is to screen all the resume
and find out the relevant ones based on job
description.
Table Of Contents
Executive Summary
1 Clarity on objectives, challenges, and desired outcomes
4
Problem Statement and Objective
2 Defined boundaries of services, deliverables, and responsibilities
5
Scope Of Work
3 Key factors and conditions required for successful execution
6-7
Dependencies
4 Clarity on objectives, challenges, and desired outcomes
8
Design
5 Blueprint of the solution aligned with business goals
9
Components And Their Description
6 Detailed breakdown of solution elements and their roles
10
Deliverables
7 Tangible outputs and outcomes committed as part of the engagement
11
Assumptions
8 Key considerations made while defining scope and execution
12
Risks and Mitigations
9 Identified challenges and strategies to minimize impact
13
Timelines and Effort Estimates
10 Planned schedule with resource allocation for each phase
14-16
3
Executive Summary
HireLens is an AI-powered resume screening system that helps
recruiters identify the most relevant candidates by analyzing
resumes against a job description. It saves time, reduces manual
effort, and increases accuracy in shortlisting. The project aims to
bring clarity, speed, and intelligence to the recruitment process.
AI-Powered Resume Screening
Automate resume shortlisting using NLP and similarity analysis to find the most
relevant candidates instantly.
Improved Recruitment Efficiency
Reduce manual screening time, eliminate repetitive tasks, and focus on the top-
qualified applicants.
Empowered Talent Acquisition Teams
Provide HR professionals with smart tools and insights to enhance decision-
making and ensure fair evaluations.
4
Problem Statement and Objective
Problem Statement Objective Statement
• Too many resumes to check manually ๏ Automate screening with AI to quickly find the
right candidates.
• Good profiles often get missed ๏ Use NLP to match skills and meaning, not just
keywords.
๏ Standardize the process for fair and unbiased
• Inconsistent shortlisting
results.
• Slow hiring process ๏ Speed up decisions with automatic ranking and
filtering.
5
Scope Of Work
Scope Of Work Feature Assumption
• ReactStores pre-fed and uploaded • All resumes are in standard
resumes in PostgreSQL or Any Database. formats.
• Supports PDF/DOCX parsing using • Parsing libraries handle all
Resume Repository Python libraries (e.g., pdfminer, python- document variations reliably.
docx). • Storage is scalable and fault-
• Indexes resumes for fast search. tolerant.
• Implements user roles (Recruiter, HR,
• Active Directory integration may
Admin) using JWT-based authentication.
Role-Based Access • exist otherwise, internal RBAC is
Controls permissions for viewing,
Control uploading, editing, or deleting resumes
used.
• Authentication system is secure.
and JDs.
• Built with ReactJS + Tailwind CSS for
recruiter interface. • Users access dashboard via
• Displays matched resumes, relevance
modern browsers.
Dashboard UI scores, highlighted skills, and experience • Backend APIs are responsive and
alignment.
• Supports shortlist, download, and return structured JSON.
messaging actions.
• Matches resumes with JDs using
embeddings (Sentence-BERT or OpenAI • Embedding models provide
embeddings). accurate semantic similarity.
AI Matching Engine 6
Scope Of Work
Scope Of Work Feature Assumption
• Extracts skills, experiences, and • Job descriptions follow standard
Job Description qualifications from uploaded JDs using NLP. format.
Parser • Supports keyword extraction, entity • NLP models are trained/fine-tuned
recognition, and semantic understanding. for domain-specific terminology.
• External HR systems expose
• REST APIs built with FastAPI/Flask. standard APIs.
Integration API • Accepts JSON payloads for resumes and JDs. • Payload structure and
authentication are compatible
7
Dependencies
•Procurement and provisioning of AI/NLP infrastructure for semantic matching, vector embeddings, and
NLP parsing.
• Assurance that all candidate and JD data remains secure, with RBAC and encryption applied for both
storage and system access.
• Engagement of domain experts or HR SMEs to define success criteria, golden datasets, key skills, and
candidate evaluation rules.
• Alignment between HireLens team and Client/Recruiter on system scope, architecture, and screening
criteria before development kickoff.
• Preparation of pre-fed resumes and job descriptions by Client in approved formats (.pdf, .docx, .txt)
with proper metadata tagging.
• Agreement on the technology stack, including ReactJS for frontend, Python (FastAPI/Flask) for backend,
PostgreSQL/MySQl for storage, and HuggingFace/OpenAI models for NLP matching.
• Finalization of Dashboard UI/UX, scoring rules, and API integration requirements during backlog/story
definition.
• Direct integration with Client HR systems or SQL databases for optional automation, ensuring resume
and document formats comply with storage and API limits.
8
Design
9
Components And Their Description
Extracts candidate details such as name, contact, skills, experience, and education from
Resume Parser uploaded resumes (PDF, DOCX, TXT).
Extracts required skills, qualifications, responsibilities, and role details from uploaded job
Job Description Parser descriptions.
Compares parsed resumes with job descriptions using AI/NLP algorithms to compute
Matching Engine relevance scores and rank candidates.
User interface for recruiters to view ranked candidates, filter results, and gain insights on
Dashboard UI skills and suitability.
Secure storage for resumes, job descriptions, processed data, and relevance scores for future
Database reference.
Analytics Module Provides insights such as top skills, experience distribution, and skill gaps among candidates.
Ensures different levels of access for admins, recruiters, and other users to protect sensitive
Role-Based Access Control data.
Resume Preprocessing Cleans and standardizes resumes to handle formatting issues, missing data, or inconsistent
Module structures before parsing.
File Upload & Storage Handles uploading resumes and job descriptions securely, with proper organization and
Module versioning.
1
0
Deliverables
AI-Based Resume Screening System – A fully functional system that parses
resumes and job descriptions, ranks candidates, and shortlists the most relevant
ones.
Recruiter Dashboard – A user-friendly interface to view ranked candidates, filter
results, and gain insights on skills and suitability.
Processed & Ranked Candidate Lists – Shortlisted candidates with relevance
scores and comparison against job requirements.
1
1
Assumptions
• Resumes and job descriptions are provided in supported formats such as PDF, DOCX,
or TXT.
• Job descriptions contain clear details about required skills, qualifications, and
responsibilities.
• Users (recruiters/admins) have basic knowledge to operate the dashboard.
• Sufficient infrastructure and server resources are available to handle AI processing
and data storage.
• Candidate data is accurate, complete, and up-to-date.
• AI/NLP models used for matching are properly trained and maintained.
• Security measures are in place to protect sensitive candidate and job data.
• Internet connectivity is stable for cloud-based processing or remote access.
• The system will be regularly updated to handle new resume formats, job description
structures, and evolving AI models.
Risks and Mitigations
Risk: Poor resume quality or inconsistent formats.
Mitigation: Implement robust preprocessing and parsing to handle formatting issues and
missing data.
Risk: High server load due to bulk uploads.
Mitigation: Optimize infrastructure, implement queueing mechanisms, and scale resources as
needed.
Risk: Incorrect or incomplete job descriptions affecting matching accuracy.
Mitigation: Provide guidelines for JD uploads and allow recruiter review of extracted data.
1
3
Timeline
Week-1 Week-3-6 Week-8
Requirement Development & Integration Deployment &
Gathering Documentation
• Collect system • Build resume parser, job description • Deploy the system,
requirements, user parser, matching engine, dashboard, and prepare user manuals, and
needs, and data formats. database integration. technical documentation.
Week-2
System Design & Week 7
Architecture Testing & QA
• Design the overall system • Perform functional testing,
blueprint and workflow. AI model validation, and bug
fixes.
1
4
Predictive Analysis Timeline
Week-1 Week-3-6 Week-8
Model Selection, Training & Testing Deployment &
Data Collection &
Preparation Monitoring
• Choose appropriate AI/NLP models for candidate-job
• Gather resumes, job matching. Train the AI models on the prepared data to • Deploy the predictive
descriptions, and predict relevance scores. Test model predictions, refine analysis module in
historical hiring data; accuracy, and validate scoring against sample data. production, monitor
clean and preprocess Connect predictive analysis module with the main system performance, and make
data. and dashboard. final adjustments.
Week-2 Week 7
Feedback Loop
Feature Extraction
Implementation
• Identify key skills, • Set up mechanism to
experience, education, and
other relevant features improve model predictions
from resumes and job based on recruiter feedback
descriptions. and hiring outcomes.
1
5
Timelines and Effort Estimates
Domain Members
React Developer (UI) 1
Backend / Database Developer 1
AI Developer 1
Tester 1
1
6
Thank You!
1
7