FRIDAY: ONLINE · OPEN TO OPPORTUNITIES

Vaibhav
Bansal

I build AI systems that ship LLMs, RAG pipelines, and full-stack products with 5+ years in production.

>
View Work Resume

featured projects

01

Disaster Response AI Copilot

02

Crypto Trade Prediction

03

Heart Disease Prediction with Generative AI

04

Car Price Prediction — MLflow + Kafka + Debezium

05

Data Science Salary Prediction Platform

06

Realtime Logs Processing — Airflow + Kafka + Elasticsearch

tap a project for details

skills × projects

live force simulation · hover a node

Initializing skill matrix...

Skills &
Expertise

From production AI systems to scalable full-stack apps — I work across the entire product lifecycle, from design through deployment.

AI / ML
LangChainRAGLLMsTensorFlow
Frontend
ReactNext.jsTailwind CSSRedux
Backend
Node.jsFlaskFastAPISpring Boot
DevOps
DockerKubernetesAWSAzure
№ 01 — Springer · VIT

Obstacle Avoidance Using Stereo Vision and Depth Maps for Visual Aid Devices

Proposed an integrated real-time obstacle detection framework for visually impaired users, combining dual CCD stereo cameras with ultrasonic sensors to produce dense depth maps via disparity estimation. The algorithm categorises 3D point vertices into near/far zones — ultrasonic handles close-range hazards (<2 m) while stereo vision covers mid-to-long range. A Raspberry Pi processes frames on-device; alerts are delivered through a buzzer, voice module, and SMS via GPS co-ordinates. The system overcomes the depth-blindness of monocular approaches and achieves reliable detection in both indoor corridors and outdoor environments. Published June 2020 in Springer Nature · SN Applied Sciences, Vol. 2, Issue 6, Article 1131. DOI: 10.1007/s42452-020-2815-z.

Year

2020

Venue

Springer Nature

Domain

Deep Learning · NLP

ORCID

0000-0002-5433-0385

Read the paperORCID profile
Computer VisionStereo VisionDepth MapsAssistive Technology