Help & Supportcareers@gravityer.com
AI Product Developer Intern with less than a year in Generative AI and RAG Systems
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Generative AI Engineer specializing in LLM fine-tuning (LoRA/QLORA), RAG system architecture, and prompt engineering. Deep experience adapting open-source models (LLaMA, Mistral) for domain-specific use cases from dataset curation and instruction tuning to GGUF quantization and low-latency deployment via Groq. Passionate about building LLMs that genuinely behave differently, not just respond differently
KLN College of Engineering
B.Tech · Artificial Intelligence and Data Science
August 1, 2022 – June 30, 2026
Rajan Matric Hr. Sec. School
Higher Secondary Education
June 1, 2020 – May 31, 2022
PIBI
AI Product Developer Intern
January 1, 2026 – Present
India
MentorAI — Fine-Tuned Persona-Driven Mentorship System
September 1, 2022 – June 1, 2026
Fine-tuned Llama-3.1-8B via QLoRA (4-bit NF4) on 1,200+ custom instruction pairs, converging training loss from 1.78 → 1.02 on a single T4 GPU. Engineered a multi-tone prompting system (Hard Push, Savage, Coach modes) and built a long-term memory layer via FastAPI + Supabase, deployed as GGUF for 8GB RAM inference.
Book-Integrated RAG System
September 1, 2022 – June 1, 2026
Built an end-to-end semantic Q&A pipeline over PDF books and YouTube videos using hybrid BM25 + semantic retrieval (70/30 weighting), dynamic adaptive chunking (500-1400 tokens), Groq Whisper for transcription, and LLAMA-3.3-70b for citation-grounded answer generation with page-level and timestamp references.
View ProjectAegis Intel Engine — Autonomous OSINT Multi-Agent Pipeline
September 1, 2022 – June 1, 2026
Engineered a full-stack autonomous intelligence system using a split 70B orchestrator + 8B writer multi-agent architecture on Groq. The 70B brain validates defense topics, coordinates regional Serper scraping, and outputs strict JSON chapter outlines; 8B agents draft 1,500+ word classified-style reports in parallel, stored in Supabase with JWT/RLS auth and dynamic .docx export.
Agentic AI Interview Simulator
September 1, 2022 – June 1, 2026
Built a voice-enabled, fullscreen AI interview system using a 5-agent LangGraph StateGraph pipeline: OSINT Insider (Tavily), Gatekeeper (chain-of-thought match scoring), Strategist (resume-aligned question planning), Adaptive Interrogator (real-time WebSocket chat with Web Speech API), and Assessor (Hire/No-Hire JSON report card). Backend on FastAPI + Groq (LLaMA-3.3-70b).
HACKHAZARDS 2025
Unknown
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's diverse range of personal projects, all centered around advanced AI applications, demonstrates a strong passion for the field and a self-driven learning attitude. This aligns well with an innovative and fast-paced AI product environment. The target role of 'AI Product Intern' is a strong match for the candidate's current experience and future aspirations, indicating a good cultural fit for a growth-oriented team.
Soft Skills & Operational Fit
The candidate's project descriptions indicate strong problem-solving skills, an ability to work with complex systems, and a proactive approach to learning and applying new technologies. The focus on building end-to-end solutions suggests good operational fit for roles requiring independent development and integration.