	In Browser
	StumbleUpon
	del.icio.us
	Google
	Google Buzz
	reddit
	LinkedIn

	Facebook
	Twitter
	Linkedin
	E-Mail

Generative AI > Google Gemini API > Recruitment Screening Agent

Recruitment Screening Agent

Author: Venkata Sudhakar

Recruitment teams at ShopMax India receive hundreds of applications for every open role. Manually reviewing each resume against job requirements is slow and inconsistent. An intelligent screening agent can score candidates objectively and surface the best-fit applicants instantly.

This tutorial builds a Gemini ADK agent that evaluates candidate profiles against a job requirement spec, scores them on skills match, experience, and education, and generates a ranked shortlist with a brief recommendation for each candidate.

The below example shows a recruitment screening agent for ShopMax India in a business context.

import asyncio
from google.adk.agents import LlmAgent
from google.adk.tools import FunctionTool
from google.adk.runners import Runner
from google.adk.sessions import InMemorySessionService
from google.genai import types

# Job requirement for open role
JOB_SPEC = {
    "title": "Senior Data Analyst",
    "required_skills": ["Python", "SQL", "Excel", "Data Visualisation"],
    "min_experience_years": 3,
    "preferred_education": "B.Tech or MBA",
    "location": "Bangalore",
}

# Candidate applications
CANDIDATES = [
    {"id": "APP-001", "name": "Neha Gupta",
     "skills": ["Python", "SQL", "Tableau", "Excel", "Machine Learning"],
     "experience_years": 5, "education": "B.Tech Computer Science",
     "location": "Bangalore"},
    {"id": "APP-002", "name": "Rohan Mehta",
     "skills": ["Excel", "SQL", "Power BI"],
     "experience_years": 2, "education": "BBA",
     "location": "Mumbai"},
    {"id": "APP-003", "name": "Divya Krishnan",
     "skills": ["Python", "SQL", "Excel", "Data Visualisation", "R"],
     "experience_years": 4, "education": "MBA Analytics",
     "location": "Bangalore"},
    {"id": "APP-004", "name": "Arjun Patel",
     "skills": ["Python", "Data Visualisation"],
     "experience_years": 3, "education": "B.Tech IT",
     "location": "Hyderabad"},
]

def screen_candidates() -> dict:
    results = []
    for c in CANDIDATES:
        # Skills match score (max 40)
        matched = [s for s in JOB_SPEC["required_skills"] if s in c["skills"]]
        skills_score = int(len(matched) / len(JOB_SPEC["required_skills"]) * 40)
        # Experience score (max 30)
        if c["experience_years"] >= JOB_SPEC["min_experience_years"] + 2:
            exp_score = 30
        elif c["experience_years"] >= JOB_SPEC["min_experience_years"]:
            exp_score = 20
        else:
            exp_score = 5
        # Education score (max 15)
        edu_score = 15 if any(e in c["education"] for e in ["B.Tech", "MBA"]) else 5
        # Location score (max 15)
        loc_score = 15 if c["location"] == JOB_SPEC["location"] else 5
        total = skills_score + exp_score + edu_score + loc_score
        results.append({
            "candidate_id": c["id"],
            "name": c["name"],
            "location": c["location"],
            "matched_skills": matched,
            "experience_years": c["experience_years"],
            "education": c["education"],
            "total_score": total,
            "recommendation": "Shortlist" if total >= 70 else ("Review" if total >= 50 else "Reject"),
        })
    results.sort(key=lambda r: r["total_score"], reverse=True)
    return {"role": JOB_SPEC["title"], "applicants": len(results), "ranked": results}

screen_tool = FunctionTool(func=screen_candidates)

agent = LlmAgent(
    name="recruitment_agent",
    model="gemini-2.0-flash",
    description="Screens job applicants for ShopMax India and produces a ranked shortlist.",
    instruction=(
        "You are a recruitment screening agent for ShopMax India. "
        "When asked to screen applicants, call screen_candidates. "
        "Present results ranked by score, grouped as Shortlist, Review, and Reject. "
        "Include matched skills, experience, and a one-line recommendation per candidate."
    ),
    tools=[screen_tool],
)

async def main():
    session_service = InMemorySessionService()
    runner = Runner(agent=agent, app_name="recruit_app", session_service=session_service)
    session = await session_service.create_session(app_name="recruit_app", user_id="hr1")
    msg = types.Content(role="user", parts=[types.Part(
        text="Screen all applicants for the Senior Data Analyst role"
    )])
    async for event in runner.run_async(user_id="hr1", session_id=session.id, new_message=msg):
        if event.is_final_response() and event.content:
            print(event.content.parts[0].text)

asyncio.run(main())

It gives the following output,

Screening Report - Senior Data Analyst | ShopMax India

Total Applicants: 4

[SHORTLIST]
  1. Neha Gupta (Bangalore)    | Score: 95 | Exp: 5 yrs | Skills: 4/4 matched
     Recommendation: Strong all-round fit. Schedule technical interview.

  2. Divya Krishnan (Bangalore)| Score: 90 | Exp: 4 yrs | Skills: 4/4 matched
     Recommendation: Excellent match with MBA Analytics background.

[REVIEW]
  3. Arjun Patel (Hyderabad)   | Score: 55 | Exp: 3 yrs | Skills: 2/4 matched
     Recommendation: Meets experience threshold but missing SQL and Excel.

[REJECT]
  4. Rohan Mehta (Mumbai)      | Score: 35 | Exp: 2 yrs | Skills: 2/4 matched
     Recommendation: Below minimum experience. Location mismatch.

ShopMax India can integrate this agent with its applicant tracking system to automatically score every incoming application. The HR team then reviews only the Shortlist tier, cutting screening time from days to minutes while reducing unconscious bias through consistent objective scoring.

Send your comments, suggestions or queries regarding this site to [email protected].