Pratham UG

+•• ••••• •••••••••••@•••••.com

1.2 yrs

Ask about Pratham

Skills

Claimed

Code ParsingData Structures and AlgorithmsSentence Transformer UsagePDF RAG System DevelopmentEnd-to-End Data Science Lifecycle ManagementTemporal Order PreservationSARIMA ModelingARIMA ModelingForecasting System DevelopmentCode Explanation GenerationPython Code Explanation GenerationRetrieval Quality Evaluation

Proof of Experience

AI Engineer

TURYON Tech Labs Pvt. Ltd. (Product: SyntaxLens) · Aug 2025 – May 2026

  • Architected the core backend of SyntaxLens, an AI-powered code-understanding system, integrating Abstract Syntax Tree (AST) parsing with LLM-driven pipelines to generate multi-level Python code explanations.
  • Designed and orchestrated complex, self-correcting agentic workflows using LangGraph and FastAPI, enabling dynamic context-retrieval and achieving a 94% reduction in model hallucinations across enterprise repositories.
  • Engineered highly scalable semantic retrieval pipelines using FAISS vector databases and sentence embeddings, reducing search latency by 32% for concurrent user queries.
  • Formulated advanced prompt engineering and context-window optimization strategies, minimizing API token utilization by 28% while maintaining deterministic, auditable outputs.
  • Developed robust, secure REST APIs using FastAPI, implementing secure routing, exception handling, and seamless integration with downstream services.
Intellipaat

Data Science Intern

Intellipaat · Feb 2025 – Aug 2025

  • Managed and executed end-to-end machine learning lifecycles, transforming raw data ingestion feeds into fully validated classification, forecasting, and NLP models.
  • Developed and deployed robust data preprocessing and feature engineering pipelines, using statistical modeling and multivariate regression to prepare high-dimensional inputs.
  • Conducted hypothesis testing and statistical analysis to validate model improvements, documenting predictive lift using p-values and confidence intervals to ensure compliance transparency.
  • Built interactive analytics dashboards in Streamlit, communicating core model performance metrics (ROC-AUC, feature importance, confusion matrices) to non-technical business stakeholders.
  • Applied MLOps best practices to establish model versioning, continuous containerized deployment via Docker, and real-time monitoring to detect performance and feature drift.

Proof-of-Work

No completed sessions yet. Anything Pratham finishes shows up here with the evidence behind it.

Proof of Education

Bachelor of Technology (B.Tech), Computer Science and Engineering

Presidency University · 2021 — 2025

A Proof CV — one profile, kept current, with the proof attached. Back to the portal