Niket Nishi

Client - Lexis Nexis Risk

• Developed AWS lambda function utilizing Guardrails and AzureOpenAI to extract entities from obituary HTML files, enhancing data extraction efficiency.
• Orchestrated AWS pipeline leveraging S3 triggers, Lambda, SQS, DLQ, and ECR, optimizing data processing and workflow automation.
• Implemented custom Google search API integration to retrieve top results, enriching data gathering capabilities.
• Refactored and integrated gen AI code, improving codebase scalability and maintainability.
• Designed and deployed multiple APIs utilizing LLMs for classification, parsing, embedding, and consolidation news article.
• Dockerized APIs and utilized docker-compose for streamlined deployment and scalability.
• Implemented PGVector plugin with PostgreSQL for efficient storage and retrieval of vector embeddings.
• Created OpenAPI Swagger documentation for each deployed API, facilitating seamless integration with frontend systems.

Client - Syndesis Health

• Established data pipeline with Snowflake stored procedures and cron jobs, ensuring real-time ancillary table updates.
• Engineered asynchronous APIs for patient data visualization using Gramex, Snowflake REST API, and Pandas.
• Designed and developed ICD10 code hierarchy algorithm, enhancing code accuracy and efficiency.
• Architected table schema for metadata storage, facilitating efficient patient data filtering.
• Leveraged Elasticsearch for smart search functionality, optimizing ICD10 code retrieval.
• Enhanced performance with full-text search implementation in MySQL server, complementing Elasticsearch.
• Conducted performance testing using Apache Benchmarking tool to ensure API concurrency and efficiency.

Client - BRE Hotels

• Set up data pipeline with stored procedures and cron jobs using Gramex for efficient data updates.
• Managed deployment across multiple environments using Azure virtual machines and NGINX reverse proxy.
• Developed APIs for granular KPI data extraction, enhancing data accessibility.
• Engineered recursive function for hierarchical data retrieval, improving data granularity and depth.

Client - Pepsico

• Established data pipeline from Snowflake to MySQL server, optimizing data flow and integration.
• Conducted complex calculations for data preparation at database level, reducing workload on API side.
• Developed custom functions in MySQL server for advanced calculations, ensuring data accuracy.
• Implemented stored procedures and schedulers for consistent data updates.

Client - Horizon Services

• Transitioned machine learning models to production-grade applications on Gramex, improving scalability and performance.
• Automated creation of pickle files for model deployment, reducing CPU overhead.
• Integrated ML models for cost optimization in marketing and labor spending.
• Utilized Snowflake as primary data source for ML model training, ensuring data accuracy and consistency.
• Coordinated with frontend team for seamless API integration.

• Implemented Google Analytics tags for comprehensive marketing analysis.
• Optimized Inventory and Warehouse management processes through workflow creation and database model design.
• Successfully migrated code base to Python 3.x and Django 3.x for enhanced performance and compatibility.
• Integrated multiple third party APIs for logistics and order status communication.
• Designed customer-centric features to enhance user experience.
• Managed user authentication and authorization, ensuring secure access.
• Automated tasks using Linux cron jobs.
• Utilized modal for data collection and utilized data for marketing initiatives.
• Deployed applications on AWS EC2 using Gunicorn and Nginx.
• Managed environment variables for efficient configuration management.

Client - Rapawalk Fashion Technologies Pvt. Ltd.

• Developed dynamic landing page features to highlight curated collections, enhancing customer engagement.
• Implemented token-based authentication and utilized Local Storage and Session Storage for efficient pop-up presentation.
• Introduced new workflows to streamline return and exchange processes, improving operational efficiency.
• Optimized image display using the Pillow library for preprocessing images of varying sizes, enhancing website performance.
• Integrated Razorpay API for generating payment links, facilitating seamless payment processing.
• Utilized Razorpay webhook for order processing post-payment via payment links, ensuring smooth transaction flow.

Client - WorthIT

• Developed customized templates for ERP modules (Sale Order, Purchase Order, Invoice, Delivery Reports) using XML to meet specific business requirements.
• Provided crucial production support by addressing and resolving bugs in the WorthIT system.
• Introduced new database models and views to align the system with evolving business needs.
• Implemented strategic modifications to base models and templates through inheritance and method overriding, ensuring seamless integration of custom solutions.
• Leveraged a technology stack comprising Python, PostgreSQL, XHTML, and Open ERP (Odoo) to drive business-driven enhancements.