ML Engineer with less than a year in deep learning, MLOps, and cloud deployment.
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Computer Science undergraduate (graduating May 2026) specializing in machine learning and AI engineering, with end-to-end experience across the ML lifecycle: fine-tuning, computer vision, LLM/RAG systems, evaluation, and production deployment. Skilled in Python and PyTorch across deep learning (YOLOv8, Mistral-7B QLORA), classical ML, RAG, MLOps, and cloud deployment on AWS and GCP.
Vellore Institute of Technology, Vellore
B.Tech · Computer Science
August 1, 2022 – June 30, 2026
Army Public School, Jalandhar Cantt
Class XII · Science
June 1, 2021 – May 31, 2021
Army Public School, Jalandhar Cantt
Class X · General
June 1, 2018 – May 31, 2019
RedoQ Software Services Pvt. Ltd.
Web Developer - Intern
May 1, 2025 – July 31, 2025
India
SentinelOps: Agentic LLM Copilot for SRE Incident Response
June 19, 2026 – Present
Constructed an end-to-end MLOps platform fine-tuning Mistral-7B-Instruct with QLORA (4-bit NF4, rank-16 LORA) on 2,500+ self-compiled postmortems over 21 GPU-hours (PyTorch, PEFT, TRL, W&B, MLflow); 0.63 faithfulness on a held-out set graded by a Llama-3.3-70B LLM-as-judge, with failure modes in a published HF model card. Built a two-stage RAG agent in LangGraph orchestrating tools (Prometheus, runbook search, postmortem drafting) over a 2,579-vector Qdrant index with BGE embeddings + a cross-encoder reranker; served via FastAPI + WebSockets with VLLM AWQ inference on Modal. Shipped a TensorFlow/Keras DistilBERT baseline and a Ragas + LLM-as-judge eval suite in a GitHub Actions CI eval-gate; deployed to Kubernetes via Helm + ArgoCD GitOps with Prometheus/Grafana observability and Kafka (Redpanda) ingestion.
View ProjectSafety Vision: Open-Source AI Workplace-Safety Monitor
June 19, 2026 – Present
Fine-tuned a YOLOv8s PPE detector (ONNX, OpenCV, CPU-only 500-800 ms) on 80k+ images with Albumentations augmentation; 0.763 mAP@50 on a held-out test set, exported for serverless inference. Built dual explainability (GradCAM + SHAP) and OSHA-grounded incident reports via a multimodal Gemini Flash LLM with RAG (Qdrant + BGE over 29 CFR standards) in LangGraph. Deployed 3 surfaces: Next.js 14/Vercel, Gradio/HF Spaces, and a serverless AWS API (Lambda Function URLs, S3, DynamoDB, ECR; Terraform; FastAPI) with a PyPI SDK + CLI; added Prophet vs SARIMA forecasting, 3 A/B tests (Cohen's d = 0.65, McNemar p < 0.001), Supabase, and a GitHub Actions CI gate (142 tests, MLflow + W&B) at $0 runtime.
View ProjectIoT Crop Yield Prediction Architecture
June 19, 2026 – Present
Deployed a custom-tuned CatBoost model on a self-compiled 250,000-row dataset (about 98% test-set accuracy) for real-time crop-yield forecasting, with a Flask service preprocessing IoT sensor data (temperature, humidity, soil nutrients) via Pandas/NumPy. Optimized the MongoDB/NoSQL schema for high-volume ingestion, cutting inference latency from a 15s baseline to about 150 ms via efficient indexing.
View ProjectAWS Certified Cloud Practitioner
Amazon
June 1, 2026 – Present
GenAI Foundation
Microsoft
June 1, 2026 – Present
DevOps Professional
Oracle
June 1, 2026 – Present
MySQL Implementation
Oracle
June 1, 2026 – Present
Data Science Professional
Oracle
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's portfolio showcases a strong initiative through multiple personal projects that are highly relevant to the ML Engineer role. The diversity of projects (LLM copilot, IoT prediction, safety vision) demonstrates a broad interest and ability to apply ML solutions to different domains. The use of open-source tools and contributions (published HF model card, PyPI SDK) suggests a collaborative and community-oriented mindset. The candidate's academic background and certifications further reinforce a commitment to continuous learning and professional development, aligning well with a growth-oriented culture.
Soft Skills & Operational Fit
The candidate's project descriptions indicate a strong ability to work on complex, end-to-end systems, suggesting good problem-solving and project management skills. The detailed technical descriptions imply clear communication of technical concepts. The breadth of technologies used across projects suggests adaptability and a willingness to learn new tools. However, without direct interview data, specific soft skills like teamwork, leadership, or stress handling cannot be fully assessed.