Help & Supportcareers@gravityer.com
AI Engineer with less than a year in Data Science, Machine Learning & Dashboard Reporting
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Final-year AI & Data Science undergrad (graduating 2026) who ships working systems — a zero-shot web data extraction pipeline powered by Mistral 7B, a real-time multilingual sentiment dashboard, end-to-end retail analytics, and an emotion-to-Spotify recommendation engine. Also builds stakeholder-ready reporting tools: real-time analytics dashboards, automated data pipelines (95% accuracy), and Power BI / Tableau reports. Comfortable from model architecture to Gradio/Streamlit deployment and from raw data to executive-ready insight. Placed 2nd at a 24-hour hackathon.
IIMT College of Engineering
B.Tech · Artificial Intelligence & Data Science
August 1, 2022 – June 30, 2026
Google Developer Group (GDG) Noida
Core Organizer - Data & AI Nexus
January 1, 2024 – December 31, 2025
Noida, Uttar Pradesh, India
RetailPulse Analytics
June 1, 2026 – Present
Built an end-to-end retail analytics pipeline covering ETL, exploratory analysis, customer segmentation, forecasting, churn prediction, and executive dashboarding using synthetic enterprise-scale retail datasets. Engineered production-style analytics workflows including RFM customer segmentation, revenue forecasting with seasonal trend modelling, churn probability scoring, and KPI reporting across products, channels, and regions. Developed SQL-based business reporting and visualisation modules to surface actionable insights – high-value customer cohorts, regional performance trends, seasonal sales spikes, and revenue optimisation opportunities for stakeholder decision-making.
Autonomous Web Data Extraction Pipeline
June 1, 2026 – Present
Built an AI-driven scraping system that takes a plain-English goal - e.g. "Get the latest AI papers from arXiv" - and autonomously determines the target URL, infers the data schema, fetches and cleans the page, and returns structured JSON records. Zero site-specific code required. Four-module pipeline: Planner (Mistral 7B via Ollama converts goal → extraction schema), Executor (httpx + BS4 strips noise), Extractor (LLM pulls structured JSON), Storage (SQLite3 + CSV/JSON export). Entire system runs locally – no API keys, no cloud, no cost. Evaluated across 5 diverse site categories: 95% JSON parse success rate, 93% increase in average records per run, 87-100% field completeness. Key innovation: zero-shot schema inference with no prior training on the target site.
Atlas - Offline LLM Assistant
June 1, 2026 – Present
Ran a quantized LLM entirely on a standard laptop – no API calls, no cloud. Implemented token streaming and short-term memory from scratch so responses feel like a real conversation, not batch output. Forces genuine understanding of LLM internals: quantization tradeoffs, context window limits, and inference latency on consumer hardware. Not a wrapper – a ground-up local AI system.
Global Sentiment Reporting Dashboard
June 1, 2026 – Present
Built a live analytics dashboard that ingests multilingual data, cleans and transforms it, runs transformer inference, and renders a colour-coded geo-map across 10+ countries — updated in real time. Designed for non-technical stakeholders: hover stats, country-level breakdowns (e.g. India: Positive 60%, Negative 25%, Neutral 15%), and a clean visual layout. Full MIS-style pipeline: ingestion → cleaning → aggregation → dashboard → business insight.
Emotion-Based Song Recommender
June 1, 2026 – Present
Trained a multi-class emotion classifier (6 categories) on text input; wired to Spotify API for live playlist generation — full loop from typed text to queued songs. Deployed as a public Streamlit app; validated on real user input across diverse writing styles, not just held-out test sets.
Business Analytics & MIS Dashboard
Deloitte Australia Simulation (Forage)
February 1, 2026 – Present
Cultural Fit Analysis
The candidate's involvement in organizing tech events (GDG Noida) and participation in hackathons suggests a proactive, community-oriented, and collaborative spirit. The diverse range of personal projects, from web data extraction to sentiment analysis and retail analytics, indicates a broad interest in applying AI/ML to various domains. The focus on building 'working systems' and 'stakeholder-ready reporting tools' aligns with a results-driven culture. However, the lack of professional experience beyond a part-time organizer role means their fit in a corporate engineering team environment is yet to be fully tested.
Soft Skills & Operational Fit
The candidate demonstrates strong initiative and organizational skills through their role as a Core Organizer for GDG Noida, managing large-scale events and coordinating with multiple stakeholders. The ability to build and deploy user-facing applications (Streamlit apps) suggests a product-oriented mindset. Their participation in a national-level sport indicates discipline and pressure handling. The project descriptions are clear and highlight practical outcomes and innovations.