Help & Supportcareers@gravityer.com
AI Engineer with less than a year in GenAI, LLM, and Machine Learning
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Highly motivated and results-driven AI Engineer with 0.7 years of hands-on experience in developing and deploying intelligent systems. Proficient in LLM Engineering, GenAI, and Machine Learning, with a strong foundation in Python, PyTorch, and cloud technologies. Proven ability to build end-to-end AI agent workflows, optimize performance, and deliver production-grade solutions, as demonstrated through successful internships and innovative projects.
SRKR Engineering College (JNTUK)
B.Tech · Artificial Intelligence & Data Science
January 1, 2022 – January 1, 2026
Narayana Junior College, Vijayawada
Pre-University Course · PCM+E
January 1, 2020 – January 1, 2022
Blackbucks Education
AI Agents Workflow Engineer Intern, Remote
December 1, 2025 – April 1, 2026
India
NIELIT
Machine Learning Intern, Remote
June 1, 2023 – July 1, 2023
India
GPT-Style Transformer Implementation & Fine-Tuning
June 1, 2026 – Present
Implemented a GPT-style decoder-only transformer from scratch in PyTorch — multi-head self-attention, causal masking, learned positional embeddings, layer norm, and feed-forward blocks — without relying on prebuilt model classes; achieved perplexity reduction of 23% over vanilla RNN baseline. Built a custom byte-pair encoding (BPE) tokenizer and training loop with gradient accumulation, LR warmup + cosine decay, and mixed-precision (FP16) training, reducing training time by 38%. Fine-tuned using LoRA/QLORA (4-bit quantization via bitsandbytes) and SFT on instruction-response pairs, cutting trainable parameters by over 90% vs. full fine-tuning while retaining 95%+ task accuracy; tracked via Weights & Biases.
Image-Based Learning Assistant
June 1, 2026 – Present
Built an end-to-end pipeline where students upload photos of handwritten problems or diagrams; OpenCV preprocessing (denoising, contrast normalization, perspective correction) improved OCR accuracy by 31% on real-world skewed phone-camera inputs. Fed extracted content into GPT-40 Vision for step-by-step explanations via LangChain orchestration with a FastAPI backend; handled edge cases including poor lighting, multi-column layouts, and mixed text-diagram inputs.
Agentic Multi-Modal RAG Pipeline
June 1, 2026 – Present
Architected a 7-agent document QA system (router → planner → validator → refiner → synthesizer → critic → self-reflection) using LangGraph state machines with dynamic replanning on retrieval-confidence failure. Fused dense vector retrieval (FAISS + ChromaDB) with cross-modal evidence aggregation across text, tables, and images; RAGAS evaluation showed +19% faithfulness and +22% context recall over single-modality RAG baselines.
Fine-Tuned EdTech QA Model
June 1, 2026 – Present
Fine-tuned Mistral-7B/Llama 3.1 8B on SciQ and ARC science-QA datasets using QLoRA (4-bit NF4 quantization) on a single GPU; used Hugging Face's PEFT to attach LoRA adapters, reducing GPU memory usage by 60% vs. full fine-tuning. Built a RAGAS-based evaluation harness measuring faithfulness, answer relevancy, and hallucination rate; achieved 21% reduction in hallucination rate and +14% RAGAS faithfulness score over the base model.
100+ LeetCode solved
LeetCode
June 1, 2026 – Present
5-Star Python
HackerRank
June 1, 2026 – Present
Computing with Python (81%)
NPTEL
June 1, 2026 – Present
GATE 2026 Dual Qualifier (Data Science & AI)
Unknown
January 1, 2026 – Present
GATE 2026 Dual Qualifier (CS & Engineering)
Unknown
January 1, 2026 – Present
Cultural Fit Analysis
The candidate's diverse personal projects, ranging from transformer implementation to multi-modal RAG and image-based assistants, indicate a strong passion for AI and a proactive learning attitude. Their internships, while limited in duration, show exposure to real-world application development and team environments. The GATE qualifications and Hackathon win suggest a competitive and achievement-oriented mindset. The focus on AI/ML aligns well with an AI Engineer role, demonstrating a strong cultural fit for a technically driven environment.
Soft Skills & Operational Fit
The candidate demonstrates strong problem-solving skills through complex project implementations and a proactive approach to optimizing performance (e.g., reducing training time, improving OCR accuracy, cutting API calls). Their experience in owning full ML lifecycles and delivering production-grade systems indicates good operational discipline and ability to work independently. The detailed project descriptions suggest strong communication of technical concepts.