NVIDIA Meets CrewAI: The Future of Automated Documentation?
November 5, 2025
AI That Writes Your Docs: My Deep Dive into CrewAI’s NVIDIA-Powered Documentation Generator
Read more →I’m a developer driven by curiosity and a love for building. I enjoy exploring new technologies, experimenting with ideas, and turning concepts into working solutions. Whether it’s diving into fresh tools, tackling challenging problems, or creating projects from scratch, I thrive on learning and making things that are both useful and exciting. Also, yes — I use Vim and Arch, btw.
Kicking Off My Portfolio Blog
September 1, 2025
An introduction to my blog, what I’ll be writing about, and how I plan to share my journey as a developer and researcher.
Read more →Things I vibe with
DevelopmentType SafetyMinimalismProductivityClean Code
Currently I am
- Working at Aveostack and AveoEarth as a founders office fullstack intern.
- Building @PaperLens, an AI powered research ecosystem.
- Experimenting with GANs for cross-domain image translation.
- Touching grass soon… probably.
Skills
PythonTypescriptPyTorchOpenCVNumPyPandasFastAPILangchainGolangApache AirflowNode.jsNext.jsPostgresSQLMongoDBGitDockerLinux
Experience
Founders Office Fullstack Intern | AveoEarth
Driving early-stage product execution by building a sustainability-focused e-commerce platform and a scalable startup services marketplace covering registration, hosting, websites, marketing, and growth tooling.
Computer Vision Intern | Delhi Technological University
Domain adaptation for optical flow in degraded weather (fog/haze) by building guided restoration pipelines, improving temporal consistency, and curating synthetic datasets for robust cross-domain training.
AI Engineer Intern | Prodigal AI
Built a high-performance hybrid retrieval pipeline (keyword + semantic search) with FAISS and LangChain, enabling accurate startup ranking and recommendations through a FastAPI–MongoDB service.
Machine Learning Research Intern | Guru Gobind Singh Indraprastha University
Built CV and speech models for candidate analysis; achieved 89% accuracy on personality prediction and 90% on speech emotion recognition.
Software Engineer Intern | Agnisys Technology Inc.
Built a high-performance sensor data visualization tool using React and Node.js, enabling smooth exploration of large JSON files via streaming, virtualization, and scalable backend storage with GridFS.