Help & Supportcareers@gravityer.com
LLM Engineer with less than a year in AI/ML engineering
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
NLP and LLM Engineer with hands-on experience building agentic systems, evaluation pipelines, and full-stack AI applications. Skilled in Python, PyTorch, HuggingFace, LangGraph, and RAG-based architectures. Seeking AI/ML engineering roles to build production-grade intelligent systems.
Dhirubhai Ambani University (DA-IICT)
M.Sc. · Data Science
July 1, 2024 – June 1, 2026
Gujarat University, Dept. of Computer Science
B.Sc. · Computer Science
October 1, 2021 – June 1, 2024
TAPS - Temporal Alignment for Personalized Summarization (Submitted to NeurIPS '26)
June 10, 2026 – Present
Designed and implemented a suite of novel evaluation metrics to quantify temporal alignment between LLM-generated summaries and evolving user preferences, using the Microsoft PENS dataset (~300K article-summary pairs). Built user alignment scores measuring preference drift over time and document faithfulness scores assessing factual consistency, enabling granular decomposition of personalization quality in LLM outputs. Developed a modular Python pipeline supporting configurable metric computation across user cohorts, document types, and temporal windows; results directly informed ICML 2026 submission findings.
Auto-SRE - Autonomous Self-Healing Agentic Framework
June 10, 2026 – Present
Architected a production-inspired multi-agent system using LangGraph to autonomously diagnose, research, and patch software defects with zero human intervention across cascading failure scenarios. Implemented a 3-node DAG pipeline: an SRE Observer capturing runtime telemetry, a Librarian Researcher performing real-time RAG via Tavily API against live documentation (eliminating knowledge-cutoff hallucinations), and a Developer Healer synthesizing code fixes. Integrated a cyclic verification loop where the SRE node re-evaluates patched code iteratively, terminating only on a clean exit code reducing MTTR on simulated production incidents. Achieved <500ms inference latency using Groq LPU-accelerated Llama 3.3 70B, enabling heavy-duty agentic reasoning without GPU infrastructure.
View ProjectJEE Gurukul - AI-Powered Adaptive Learning Platform
June 10, 2026 – Present
Built an end-to-end AI-driven JEE preparation platform leveraging Google Gemini for NCERT-aligned question generation with dynamic topic selection orchestrated via Google Cloud Platform. Implemented semantic retrieval using FAISS over a vector index of NCERT concepts, ensuring generated questions align with curriculum scope and difficulty distribution. Engineered a full-stack web application (Flask, PostgreSQL) with Google OAuth, adaptive difficulty testing, real-time quiz feedback, mobile-first UI, and per-user progress tracking across sessions. Selected as Semi-Finalist at the 100x Engineers Buildathon 2.0 (Bangalore) - top-tier national AI product competition.
View ProjectICML 2026 Submission
ICML
January 1, 2026 – Present
Semi-Finalist, 100x Engineers Buildathon 2.0 (Bangalore, 2024)
100x Engineers Buildathon 2.0
January 1, 2024 – Present
Cultural Fit Analysis
The candidate's project portfolio showcases a strong alignment with an LLM Engineer role, covering research, agentic systems, and full-stack AI applications. The diversity of projects (academic, personal, competition) indicates a broad interest and ability to adapt to different problem domains. Their involvement in a national AI product competition and a research submission suggests a drive for innovation and excellence, which aligns well with a dynamic technical environment.
Soft Skills & Operational Fit
The candidate demonstrates strong initiative and problem-solving skills through their project work, particularly in architecting autonomous systems and developing novel evaluation metrics. Their participation in a buildathon and research submission indicates a proactive and collaborative approach. The detailed project descriptions suggest good communication of technical concepts.