Headline

Applied AI/Deep Learning engineer

About

Hello! I’m Ketan Palankar, I build deep learning/AI systems that move from research to production, at scale.I’m a Lead Data Scientist with 7+ years of experience across Computer Vision, Digital Identity, Fraud Detection, Recommendation, Graph ML, Information Retrieval, multimodal systems, ML, Database and AI Infrastructure. At Airtel, I build B2B AI solutions using fine-tuned, lightweight models across Computer Vision and NLP, alongside D2C computer vision applications operating at population scale. My work spans the full model lifecycle: from model development and optimisation to GPU inference, distributed systems and production deployment. Previously at Quantiphi, I worked on cloud-native ML solutions and MLOps on AWS, building scalable training and deployment pipelines using services including SageMaker, EMR, Glue, ECS and S3 for enterprise clients. My engineering focus is on making AI practical: PyTorch, Transformers, GPU inference, Triton, Elasticsearch, Neo4j, Kubernetes, distributed pipelines and MLOps — with an emphasis on latency, scale, reliability and measurable business impact. More recently, I’ve been exploring multimodal reasoning, video understanding and lightweight vision-language models, alongside personal projects in accessible AI and vision-language systems. Languages: English, Hindi, Konkani, Dutch (learning) Passionate about everything tech, sports, yoga and books. That will be all. Thank you for your time!

Location

Karnataka

AI stack

hugging face

Work experience

Sr. Manager - Data Science at Airtel Digital, Jul 2024 - Sep 2026 · 2 yrs 3 mos, Bengaluru, Face Authentication Suite: Delivered a face recognition and liveliness model, along with real-time face matching and defraud platforms for D2C channels such as Customer Acquisition, Airtel Works, and B2B teams. Safety Suite: Developed a comprehensive smart safety app incorporating safety equipment object detection models, crash detection, and speed estimation using traditional ML and signal processing for on-field service workers,thereby reducing safety accident rate with efficient tracking. Multiple POCs: Executed various proofs of concept across domains, including GAN's and diffusion models for document digitization, PII /name matching using feature engineering, phonetics features, and tree based Traditional ML Modeling.

Senior Data Scientist at Airtel Digital, Jun 2024 - Jun 2024 · 1 mo, Bengaluru, Karnataka, India

Data Scientist at Airtel Digital, Aug 2021 - Jun 2024 · 2 yrs 11 mos, Bengaluru, Karnataka, India

Machine Learning Engineer at Quantiphi, Jun 2019 - Jul 2021 · 2 yrs 2 mos, Bengaluru Area, India, Hands on experience as a MLE with projects on • BFSI Domain: Tree based binary classification solution on Vehicle Fleet Management Insurance claim use case- Approval/rejection of claims using Random Forest model. • Multiple POC's on Document processing(OCR,Key-value pairs etc) ,ED diagram processing etc, Custom MLOps on AWS Sagemaker

Education

B.Tech, Electrical and Electronics Engineering, National Institute of Technology Karnataka