Karthikeyan B

AI Researcher

Bengaluru, Karnataka, India23 yrs 1 mo experience
Highly Stable

Key Highlights

  • Designed RAG-powered intelligence layer for enhanced AI recommendations.
  • Achieved 97.6% accuracy in employee retention prediction model.
  • Reduced editorial workload by 40% with BERT-based news classifier.
Stackforce AI infers this person is a Data Scientist specializing in AI solutions for diverse industries including Media, Insurance, and Telecom.

Contact

Skills

Core Skills

Machine LearningData AnalysisNlpSqlBig Data

Other Skills

PythonPandas (Software)TableauPL/SQLOracle SQL DeveloperTeradataEDAApplied Machine LearningStreamlitMLOpsData validation control (DVC)Docker ProductsKubernetesGitNatural Language Processing (NLP)

About

Hi, I’m Karthikeyan, GenAI & Agentic AI - oriented Data Scientist focused on designing intelligent systems that don’t just predict outcomes, but reason to drive measurable business impact. 🎯 My Journey: With over a decade of experience in QA Automation and Data Architecture, I bring a strong foundation in scalable systems, data reliability, and enterprise-grade design. This background naturally evolved into Data Science and AI, where I now focus on building autonomous, decision-supporting AI solutions. My work spans Machine Learning, NLP, Predictive Modeling, and Generative AI, with an increasing emphasis on Agentic AI workflows - systems that combine models, tools, and feedback loops to automate complex business decisions. 🏆 Key Achievements: ❇️ Designed and implemented a RAG- and Knowledge Graph–powered intelligence layer, modeling complex relationships across structured and unstructured data to enhance semantic search, reasoning, and explainability in AI-driven recommendations. ❇️ Developed a BERT-based news classifier (66% accuracy) that cut editorial workload by 40% and reduced content turnaround time. ❇️ Created an Employee Retention Prediction model (97.6% accuracy) that helped HR teams reduce attrition risk and act proactively. ❇️ Delivered Telecom demographic insights across 31K+ users, improving campaign precision by 25% through targeted recommendations. ❇️ Built a Cross-Selling Model (AUC = 0.85, 71% recall) enabling marketing teams to target high-potential insurance buyers effectively. 🧠 Key Skills & Expertise: Gen AI | Agentic AI | Python | SQL | Machine Learning | Deep Learning | NLP | Predictive Analytics | Data Visualization | Big Data (Hadoop, Spark) | Statistical Modelling | AWS (EC2, S3) | MLflow | DVC | Docker | FastAPI | Kubernetes (k3s) | GitHub Actions | Prometheus | Grafana | Business Problem Solving 🚀 My Unique Value: I combine over 20 years of data-driven architecture experience with a strong AI and analytics mindset, enabling me to bridge the gap between data engineering and business intelligence. My approach ensures that every ML model I build directly ties to business KPIs and operational efficiency. 🌱 Future Goals: I aim to advance into a AI Solution Architect role where I can lead impactful projects, mentor junior data professionals, and drive enterprise-level digital transformation through data. 📩 Let’s Connect: If you’re looking for someone who can turn complex business challenges into intelligent, autonomous AI solutions - reach out at b_karthii@yahoo.com or connect with me here on LinkedIn.

Experience

23 yrs 1 mo
Total Experience
3 yrs 7 mos
Average Tenure
1 yr 5 mos
Current Experience

Career break

Career transition

Dec 2024Present · 1 yr 5 mos · Bengaluru, Karnataka, India

Self-employed

Data Science & AI Professional

Jan 2024Present · 2 yrs 4 mos · Bengaluru, Karnataka, India · Hybrid

  • Key Projects & Achievements:
  • 📰 Article Classification using BERT (NLP Project)
  • ✅ Built an NLP-driven text classification model using BERT, achieving 64.9% validation accuracy, outperforming LSTM and GRU baselines.
  • ✅ Automated manual news tagging, reducing editorial effort by 40% and cutting publishing turnaround time significantly.
  • ✅ Developed a recommendation system to suggest contextually relevant articles, improving content discovery and user engagement.
  • 🚗 Vehicle Insurance Cross-Selling Model
  • ✅ Designed and trained Random Forest and XGBoost models to predict health insurance customers likely to buy vehicle insurance.
  • ✅ XGBoost achieved AUC = 0.85 and 71% recall, leading to more precise cross-sell targeting and higher conversion rates.
  • ✅ Identified key demographic and behavioral factors influencing purchase intent, enabling data-driven marketing segmentation.
  • 👩‍💼 Employee Retention Prediction
  • ✅ Built an ensemble-based attrition prediction model (Stacking Classifier) achieving 97.6% accuracy, replacing manual exit-interview reliance.
  • ✅ Conducted feature importance analysis to uncover key attrition drivers (e.g., satisfaction, project load, tenure).
  • ✅ Enabled HR teams to take proactive retention actions, reducing potential attrition-related costs by 15–20%.
  • 📊 Telecom User Demographics Dashboard
  • ✅ Processed and analyzed 31,742 mobile user records, performing EDA and feature engineering for demographic and behavioral insights.
  • ✅ Discovered top-performing regions (Maharashtra 19%, Karnataka 15%) and dominant device brands (Xiaomi 63%).
  • ✅ Delivered state-specific and age-segmented marketing insights, improving campaign precision by 20–25%.
Machine LearningData AnalysisSQLPythonPandas (Software)Tableau

Cognizant

Manager - Projects | Data Science

Feb 2021Dec 2024 · 3 yrs 10 mos · Bangalore Urban, Karnataka, India · Hybrid

Cognizant technology solutions

2 roles

Manager - Projects | QA Automation Architect

Promoted

Aug 2015Feb 2021 · 5 yrs 6 mos · Bengaluru, Karnataka, India · On-site

  • ✅ Received the Global Project Of the Year Award - Small Engagement Category
  • End-to-End Test Automation Framework:
  • ✅ Designed and implemented a comprehensive automation framework using Jubula, Shell, and Python, enabling automated regression and integration testing across Database tools - achieving ~85% test coverage automation and reducing QA cycle time by ~70%.
  • Performance, Security & Reliability Testing:
  • ✅ Defined and executed a multi-dimensional test strategy covering functional, performance, memory leak, load/stress, and security testing - resulting in 30% faster defect detection and improved product stability across customer environments.
Big DataSQLPL/SQLOracle SQL DeveloperTeradata

Senior Software Engineer

Nov 2010Jan 2015 · 4 yrs 2 mos · Bengaluru, Karnataka, India · On-site

  • End-to-End QA Strategy & Governance:
  • ✅ Led QA delivery across enterprise-scale data and reporting systems (ETM, Loblaw, AMD), defining test strategy, estimation, planning, and execution frameworks - achieving ~95% on-time delivery and zero critical defects post-deployment across multiple business domains.
  • Cross-Functional Leadership & Delivery Management:
  • ✅ Managed distributed QA teams (onsite/offshore) across projects, driving daily execution metrics, defect trend analysis, and stakeholder reporting, improving QA visibility and accountability - leading to a 25% increase in test productivity and enhanced stakeholder satisfaction.

Hp

Information Technology Developer

Aug 2004Oct 2010 · 6 yrs 2 mos · Bengaluru, Karnataka, India · On-site

  • ✅ Led ETL and Business Intelligence testing/development using Informatica, PL/SQL, and Business Objects, improving data accuracy and enabling reliable project performance reporting.
  • ✅ Implemented test automation using Silk Test, reducing manual QA effort by ~60% and accelerating regression cycles across workflow and data validation processes.
  • ✅ Performed end-to-end database and functional testing in Oracle and HP-UX environments, ensuring seamless data integration and system reliability.
  • ✅ Supported enterprise-wide process standardization through HP’s global Workflow and Portfolio Management platforms, enhancing delivery efficiency and customer experience

Eximsoft technologies pvt ltd

Member Technical

Oct 2003Aug 2004 · 10 mos · Bengaluru, Karnataka, India · On-site

  • ✅ Performed end-to-end functional and database testing for B2B leasing web services, ensuring accurate data replication, synchronization, and trigger functionality across DB2 environments.
  • ✅ Designed and executed manual test cases and managed defects using Rational Test Manager and ClearQuest, improving test traceability and defect resolution efficiency.
  • ✅ Validated seamless integration between order entry, fulfillment, and invoicing processes, ensuring data accuracy and reliability for enterprise financial workflows.

Netgalactic internet solutions pvt. ltd.

Quality Control Engineer

Apr 2003Oct 2003 · 6 mos · Bengaluru, Karnataka, India · On-site

  • ✅ Designed and executed manual test plans and test cases covering functional, integration, and regression testing for a healthcare tablet application with multi-input modes (Ink, Speech, Text).
  • ✅ Validated speech-to-text conversion, handwriting recognition, and multimedia integration to ensure accurate and reliable user interactions for medical professionals.
  • ✅ Conducted end-to-end module integration testing across communication and productivity tools (Email, Fax, Scheduler, Notes) to ensure seamless interoperability.
  • ✅ Reported and tracked defects effectively, improving application stability and usability prior to deployment.

Tsunami infosolutions

Test Engineer

Aug 2002Apr 2003 · 8 mos · Bengaluru, Karnataka, India · On-site

  • ✅ Designed and executed manual test cases and test data to validate end-to-end functionality across core modules - Sales, Service, Finance, and Warranty - ensuring smooth dealer–manufacturer interactions.
  • ✅ Performed integration and regression testing to verify seamless data flow between dealer and central systems, enhancing system reliability and transaction accuracy.
  • ✅ Identified, documented, and tracked defects through systematic bug reporting, improving overall product quality and stability before deployment.

Education

Accredian

Post Graduate Program (PGP) — Data Science & Artificial Intelligence

Jan 2024Oct 2025

University of Madras

Bachelor of Science — Computer Science

Jan 1997Jan 2000

Stackforce found 100+ more professionals with Machine Learning & Data Analysis

Explore similar profiles based on matching skills and experience