Divyansh P.

Key Skills

Software

Google Cloud Vertex AI

Other

Top Subject Matter

AI agents and LLM/creative editing orchestration for DOOH advertising platform

Autonomous agent pipelines with model-generated SQL validation and streaming LLM responses

Top Data Types

Text

Top Task Types

Bounding Box

Polygon

Segmentation

Classification

Cuboid

Polyline

Freelancer Overview

Software Engineer Intern — multi-agent AI backend optimization and Vertex AI Imagen integration (Screenify Technologies . Brings 2+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include Google Cloud Vertex AI and Other. Education includes Bachelor of Technology, Rajiv Gandhi Institute of Petroleum Technology (2027). AI-training focus includes data types such as Computer Code and Programming and labeling workflows including Evaluation, Rating, and Computer Programming.

Labeling Experience

Software Engineer Intern — multi-agent AI backend optimization and Vertex AI Imagen integration (Screenify Technologies Pvt. Ltd.)

Google Cloud Vertex AI

Built and optimized an asynchronous multi-agent backend that improved AI agent latency from ~15s to 2–3s by persisting agent outputs. Integrated Vertex AI Imagen for creative editing within an authenticated (JWT) and role-based access control (RBAC) backend. Supported iterative improvements through code review and secure backend design practices. • Persisted multi-agent outputs to Firestore for downstream evaluation and use. • Designed backend services using OOP and secure coding for an ad platform. • Integrated Vertex AI Imagen creative editing with access controls. • Deployed Dockerized services on Google Cloud and contributed in an Agile workflow.

2026 - Present

Software Engineering Intern — agent pipeline development with validation of model outputs (Orangecat Technologies)

Other

Developed REST APIs for an autonomous data-science agent and added a Tree-sitter validation layer to sanitize model-generated SQL. Debugged errors across the data/pipeline workflow to ensure reliable agent execution. Implemented multi-agent integration along with a React streaming interface for real-time LLM responses and charts. • Built 10+ Python-scripted REST APIs for agent-driven data-science tasks. • Validated and sanitized SQL produced by models using Tree-sitter. • Debugged and fixed pipeline errors across the end-to-end workflow. • Built a React streaming frontend and added unit tests for robustness.

2025 - 2025

Education

Work History