Bismay Nayak profile photo 1 on Tal

Bismay Nayak

Founder @Rola Labs | Building @TasteMap

Bengaluru

6+ years of work experience

TasteMap
Mar 20267 mosPresent

Rola Labs
Apr 20251 yr 6 mosPresent

Remote

platform.ai
Apr 20204 yrs 5 mosAug 2024

San Francisco, California, United States

Led end-to-end development of real-world AI systems across healthcare, food-tech, and fashion domains. • Built scalable ML pipelines and backend infrastructure deployed on Google Cloud (Cloud Run, Functions) • Delivered real-time implant detection (85%+ accuracy) and food doneness classification (90%+ accuracy) • Implemented self-supervised learning (BYOL, SimCLR), zero-shot search using CLIP, and generative AI (diffusion models, U-Net) • Optimized infrastructure, reducing latency and server costs through GCP service migration • Mentored junior ML fellows; led research on noisy-label training and robust loss functions • Collaborated with product and engineering teams to deliver analytics dashboards, case-sharing flows, and user feedback loops Tools & Technologies: Python, ReactJS, PyTorch, Flask, FastAPI, GCP, Docker, Kubernetes, CI/CD
Fellowship.AI
Jan 20203 mosMar 2020

• Built a CNN-based wound classification model trained on a dataset of 1,000+ images (78% accuracy) • Led data collection, cleaning, and preprocessing to ensure class balance and image quality • Researched noise-robust training using CoTeaching, Symmetric Cross Entropy (SCE), and LID-Paced Loss • Delivered analysis and benchmarks on robust model training under weak labeling conditions Focus Areas: Deep Learning, Computer Vision, Noisy Label Learning, PyTorch, Data Preparation

Bismay’s AI Stack

antigravity
antigravity
gemini
gemini
hugging face
hugging face
langchain
langchain
mcp
mcp
obsidian
obsidian

Education

Veer Surendra Sai University Of Technology ( Formerly UCE ), Burla

Bachelor of Technology - BTech, Computer Science

Veer Surendra Sai University Of Technology ( Formerly UCE ), Burla

20154 yrs2019

Join Bismay and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Bismay Nayak on Tal

Headline

Founder @Rola Labs | Building @TasteMap

About

Building toward a future where physical spaces are as accessible, interactive, and adaptive as digital ones. Interested in creating AI-driven systems that help people navigate, connect with, and experience the real world more naturally. Currently exploring generative AI, agentic systems, and human-centered product experiences.

Location

Bengaluru

AI stack

antigravity

gemini

hugging face

langchain

mcp

obsidian

Work experience

Founder at TasteMap, Mar 2026 - Present · 7 mos

Co-founder · AI & Product Lead at Rola Labs, Apr 2025 - Present · 1 yr 6 mos, Remote

Full Stack | Deep Learning Engineer at platform.ai, Apr 2020 - Aug 2024 · 4 yrs 5 mos, San Francisco, California, United States, Led end-to-end development of real-world AI systems across healthcare, food-tech, and fashion domains. • Built scalable ML pipelines and backend infrastructure deployed on Google Cloud (Cloud Run, Functions) • Delivered real-time implant detection (85%+ accuracy) and food doneness classification (90%+ accuracy) • Implemented self-supervised learning (BYOL, SimCLR), zero-shot search using CLIP, and generative AI (diffusion models, U-Net) • Optimized infrastructure, reducing latency and server costs through GCP service migration • Mentored junior ML fellows; led research on noisy-label training and robust loss functions • Collaborated with product and engineering teams to deliver analytics dashboards, case-sharing flows, and user feedback loops Tools & Technologies: Python, ReactJS, PyTorch, Flask, FastAPI, GCP, Docker, Kubernetes, CI/CD

Machine Learning Fellow at Fellowship.AI, Jan 2020 - Mar 2020 · 3 mos, • Built a CNN-based wound classification model trained on a dataset of 1,000+ images (78% accuracy) • Led data collection, cleaning, and preprocessing to ensure class balance and image quality • Researched noise-robust training using CoTeaching, Symmetric Cross Entropy (SCE), and LID-Paced Loss • Delivered analysis and benchmarks on robust model training under weak labeling conditions Focus Areas: Deep Learning, Computer Vision, Noisy Label Learning, PyTorch, Data Preparation

Education

Bachelor of Technology - BTech, Computer Science, Veer Surendra Sai University Of Technology ( Formerly UCE ), Burla

Bismay Nayak - Founder in Bengaluru | Tal