Experience

• 

  Wonder (askwonder.com) 🤔

  Mar 2019 - Nov 2019

  Sourcer
• 

  Codemill

  Oct 2020 - Dec 2020

  Ai researcher

  * Worked closely with 3D design team of Codemill Company to write a case study on the use of AI techniques in the company’s modeling software.* Led the group as Team Leader to meet weekly targets and project management.

• 

  Freelance (self employed)

  Feb 2021 - Jun 2021

  Nlp engineer

  - Improvised and debugged existing Named Entity Recognition project of the client with backend integration on FastAPI.- Deployed NER model using docker on client's internal server.

• 

  Fusemachines

  Jul 2021 - Nov 2022

  Machine learning engineer

  - Implemented Vision Transformer model and custom EfficientNet model to classify real and fake bags based on certain AOIs.- Implement Handwritten TextRecognition using CNN and RNN model for hand-written text images in English and Nepali Languages.- Improved accuracy by 6.9% in Language Classifier model for text images by implementing VGG16 and EfficientNet base models and image augmentation.- Worked on Image Alignment problem for distorted image based on template's layout using OpenCV.- Deployed ML models to production using Docker and Kubernetes. Show less

• 

  Deerwalk institute of technology

  Oct 2021 - Dec 2022

  Teaching assistant

  - Teaching Python Programming and Data Science course to students from second year.- Devised the syllabus for Web Development Course and Python Programming course for first andsecond year students under the supervision of faculty member.- Supervised and motivated clubs and tech community inside the college to conduct events.

• 

  Gritfeat solutions

  Nov 2022 - Mar 2024

  - Orchestrated a dynamic team to develop a ChatGPT-based salesman persona for RAG-based sales pipeline optimization, utilizing langchain, AzureOpenAI, and Pinecone.- Implemented and deployed RAG-based wiki and internal bots using GPT-3.5, Langchain, LLaMA Index, ChromaDB and Docker.- Fine-tuned Huggingface’s Sentence Transformer model for a scoring system in thenon-English language, optimizing with various pre-trained weights.- Created the end-to-end ML pipeline from new data collection, feature storing, training, inferencing, and deploying the model using MLflow’s Model serving in Databricks platform. Show less

  • Machine Learning Engineer - Level III

    Sept 2023 - Mar 2024

  • Machine Learning Engineer - Level II

    Nov 2022 - Sept 2023
• 

  Founders led studio

  Mar 2024 - now

  Machine learning engineer