Abhishek Kumar

Results-driven Backend Software Developer with nearly 5 years of experience in creating innovative solutions using Python, Django, and Docker. Proven track record of enhancing performance metrics and resolving legacy issues through critical thinking and problem-solving skills. Committed to delivering high-quality software solutions that drive industry advancements.

AI-Driven Document Extraction Pipeline 

• Developed an AI-powered pipeline to extract data from structured and semi-structured documents (e.g., driving licenses, certificates) with high accuracy.
• Designed and implemented ML models for image embedding (CLIP) and layout detection (YOLO).
• Built a scalable MLOps pipeline, ensuring automation, reproducibility, and deployment efficiency.
• Deployed models on Azure Cloud using Terraform for infrastructure as code (IaC).
• Implemented CI/CD workflows using GitHub Actions for automated testing and deployment.
• Secured deployment by integrating Azure Key Vault for API keys, credentials, and secrets management.

Brightstream  Enterprise Search Product

• Led MLOps and backend development for Brightstream, an enterprise search solution leveraging Retrieval-Augmented Generation (RAG), OpenAI, Langfuse, and FastAPI.
• Designed and implemented search and indexing pipelines to enhance information retrieval efficiency.
• Deployed and managed cloud infrastructure on Azure using Terraform.
• Collaborated with frontend teams to integrate search capabilities seamlessly into the product.

EdGE Jscore (Mobility) 

Python, RMQ, Docker, Elasticsearch, Pandas, Redis, SQL

• Developed an AI-driven workforce optimization tool to match employees with internal job openings based on skills and expertise, enhancing talent allocation and reducing costs.
• Integrated AI search capabilities to accelerate hiring and improve hire quality.
• Managed end-to-end internal mobility applications for Wipro & HCL, acting as the Single Point of Contact (SPOC) for clients.
• Handled Change Ticket Approval Requests (CTAR) via Jira, managing modifications, testing, and deployments across QA and production environments.

Swimming Strong

Python, Django, Docker, Pandas, NumPy, PostgreSQL, SQL, AWS

• Built a data-driven web application using ML models to analyze swimmers' performance gaps compared to top athletes
• Developed forecasting models for performance prediction and report generation, enabling detailed tracking by coaches
• Led end-to-end development, including project management and backend implementation in Python.
• Formulated action plans to address business problems and enhance system functionality.