Leading a software development team to create a natural language system that enables blind and visually impaired (BVI) users to perform advanced data analysis in STEM fields without requiring coding skills.

Designed and deployed a custom AI agent capable of executing data analysis tasks through natural language commands, achieving 85% task accuracy in initial testing.

Developed the system as a full-stack web application, containerized using Docker, and deployed on AWS for scalability and accessibility.

Ensured real-world usability by integrating cloud deployment for easy access and scalability.

Core member of the HappyFox AI team, building AI-driven features for a scalable SaaS helpdesk using Django, collaborating with cross-functional teams, utilizing continuous deployment (CI/CD) pipelines, and leveraging Git for workflows.

Designed and deployed the AI-Answer feature, an advanced Retrieval-Augmented Generation (RAG) system, that autonomously resolves user queries using knowledge base articles and historical ticket data, resulting in a 45% increase in ticket deflection and significantly reducing the need for human agent intervention.

Developed an AI-powered KB Draft Recommendation system that automatically analyzes repetitive tickets to generate draft KB articles, accelerating publication time by 40% and reducing recurring queries by 30%.

Enhanced Search and Retrieval Systems by integrating semantic search with AWS Elasticsearch and text embeddings, boosting response accuracy and search efficiency.

Boosted Agent Productivity by 70% through the AI Auto-Complete feature. Fine-tuned a large language model (LLM) to provide real-time, context-aware text suggestions in ticket replies, significantly reducing agent typing time.

Developed an internal web app for seamless data migration from other helpdesks to HappyFox using Django, REST APIs, Celery, Redis, and PostgreSQL. Improved migration reliability and reduced time by implementing efficient background tasks.

Built an internal application for analyzing subscription data, using ClickHouse to process large volumes of data and extract key metrics like Monthly Recurring Revenue (MRR) from HappyFox’s payment gateway.

Based on my performance, I was converted from an intern to a full-time Backend Engineer, further contributing to HappyFox’s AI-driven features and SaaS solutions.

Collaborated with the quants team to convert trading strategies into executable algorithms, enabling automated execution based on user-defined cron jobs. This improved execution speed by 90% and boosted profits by 50%.

Worked directly with clients to adjust and fine-tune trading strategies, adapting to their dynamic needs and proving an increase in profit through tailored solutions.

Led the development of the algorithmic trading platform, including system design, database architecture, and development using Django Rest Framework and ReactJS. Ensured efficient execution and real-time trade alerts for automated strategies.

Integrated push notifications into the platform to send real-time trade alerts, ensuring that clients stay informed of critical trading events promptly.