Conversational AI
Assistants built for the kind of answer you need.
Guided rule-based bots, meaning-aware semantic search, natural LLM assistants, database questions with Text-to-SQL, and document-grounded answers with RAG.
Hello, I am Syed Usman, an AI/ML Engineer and Full-Stack Developer from Bangalore, building assistants and computer vision products that make useful decisions easier.
Same technical depth, translated into outcomes your team can actually act on.
Conversational AI
Guided rule-based bots, meaning-aware semantic search, natural LLM assistants, database questions with Text-to-SQL, and document-grounded answers with RAG.
Computer Vision
Detection, tracking, pose estimation, fatigue signals, and movement analysis.
Structured Data
Text-to-SQL analytics for teams who need insight without writing queries.
Unstructured Data
RAG-based assistants that retrieve evidence before they answer.
For Engineers
Projects across safety, enterprise workflows, education, sport, and analytics.
Industry Client Project | E2E Rosh IT Solutions
Built computer vision modules for safety monitoring in an industrial environment: driver fatigue detection, worker proximity monitoring, oil leak detection, and pose-based fall alerts.
Full-Stack LLM | RAG & Text-to-SQL | Live Demo
A production-ready assistant merging document-grounded RAG with automated Text-to-SQL analytics. Uses intelligent query routing to direct prompts to vector stores or relational databases, returning dynamic business visualizations and structured execution details.
Python | Frappe | LLM | RAG
Plain-English HR, inventory, and finance queries inside ERPNext, designed within sandbox constraints.
Python | SQL Server | LLM
Text-to-SQL assistant producing multi-table analytics and visual reports without manual SQL.
RAG | Node.js | MongoDB
Syllabus-specific assistant with notes retrieval, authentication, uploads, and moderation.
Flutter | TensorFlow Lite | MoveNet
Mobile pose-analysis app for rep counting, form feedback, and sports performance assessment.
Computer Vision + Chatbot Intelligence
A portfolio concept combining my computer vision and assistant work: upload footage, detect players and the ball, track movement and velocity, then ask questions about the match in plain English.
I build across machine learning and product development: from pose estimation and real-time APIs to RAG pipelines and web experiences. I am currently completing my B.E. in Artificial Intelligence & Machine Learning at T John Institute of Technology.
Experience
E2E Rosh IT Solutions
January 2026 - PresentEducation
T John Institute of Technology
2022 - 2026 | CGPA: 8.0 / 10Certifications
Open To Work | Freelance Projects
I am available for AI/ML roles and projects involving assistants, data, or computer vision.