Help & Supportcareers@gravityer.com
AI Engineer with less than a year in Agentic AI systems, LLM fine-tuning, RAG pipelines, and end-to-
AI is analyzing your overall score…
Identifying your key strengths…
Evaluating your skill match against the job requirements…
Assessing your cultural and operational fit
Production-focused AI/ML Engineer specialising in Agentic AI systems, LLM fine-tuning, RAG pipelines, and end-to-end MLOps. Built multi-agent orchestration frameworks (LangGraph), multimodal retrieval systems, and low-latency FastAPI inference services on GCP/AWS. Reproduced GPT-2 (124M) from scratch achieving 11x training throughput via Flash Attention and CUDA kernel optimisation. Strong in PyTorch, Transformers, NLP, and production deep learning systems. GATE DA 2026 Qualified 300+ DSA.
Dr. A.P.J. Abdul Kalam Technical University (AKTU)
B.Tech · Computer Science & Engineering (AI & ML Specialisation)
August 1, 2022 – May 1, 2026
NexusStream: Distributed Cross-Device Multi-Agent AI Orchestrator
June 1, 2024 – June 1, 2026
Architected a single-pass LangGraph agentic core consolidating NLP classification, expert routing, and self-criticism into one execution cycle - reducing local 7B LLM latency by 66% through pipeline-level optimisation. Built a full-duplex FastAPI + WebSocket streaming backend broadcasting real-time AI outputs to cross-device clients; implemented async concurrency controls eliminating race conditions under high-frequency event streams.
View ProjectJARVIS: Multi-Tenant Distributed Autonomous Agent Platform
June 1, 2024 – June 1, 2026
Engineered a multi-tenant agentic orchestration layer using LangGraph and LiteLLM with concurrent browser automation via CDP; built a Semantic DOM Parser coupling LLM reasoners with VLMs to eliminate hardcoded selector fragility. Implemented an on-the-fly contextual RAG memory engine using PyMuPDF to ingest unstructured PDFs and compile user context into prompt windows for real-time asset generation with zero-trust data privacy.
View ProjectAutomated Sales Forecasting - End-to-End MLOps Pipeline
June 1, 2024 – June 1, 2026
Orchestrated full MLOps pipeline with Airflow DAGs and batch inference achieving 95% MAPE accuracy, zero manual intervention; managed model lifecycle via MLflow + DVC with real-time Grafana observability and regression test suite.
View ProjectGPT-2 (124M) Production Training & Kernel Optimisation
June 1, 2024 – June 1, 2026
Reproduced GPT-2 (124M) Transformer from scratch; resolved memory/compute bottlenecks via Flash Attention, fused CUDA kernels, and BF16 AMP - boosting training throughput 11x to 2,743 tok/sec on NVIDIA T4 with loss 10.94 → 0.75. Outperformed original GPT-2 checkpoint on HellaSwag benchmark through systematic kernel-level optimisation; built production-grade mixed-precision training pipeline with scalable configuration.
View ProjectAutomated US Visa Approval - Real-Time ML Inference Microservice
June 1, 2024 – June 1, 2026
Engineered REST API inference microservice at <1 ms latency, 96.8% accuracy, 7x faster than baseline; deployed via Kubernetes HPA + CI/CD; integrated Evidently AI for automated model evaluation and drift detection in production.
View ProjectMulti-Agent Reasoning System (LangGraph + RAG + GCP)
June 1, 2024 – June 1, 2026
Designed cyclic multi-agent workflow with autonomous tool-use; deployed FastAPI backend on GCP with WebSockets; implemented Multimodal RAG pipeline on MongoDB Atlas achieving 0.9 Answer Relevance / 0.65 Context Precision via LLM-as-judge eval.
View ProjectGATE DA 2026 Qualified - Data Science & AI (National-level; validates ML, stats, analytical reasoning)
Unknown
June 1, 2026 – Present
300+ LeetCode - DSA: Arrays, Trees, Graphs, DP, Hashing (Python & Java)
LeetCode
June 1, 2026 – Present
Cultural Fit Analysis
The candidate's project portfolio demonstrates a strong alignment with cutting-edge AI/ML development, particularly in agentic systems and MLOps, which are highly relevant to an AI Engineer role. The diversity of projects, from distributed orchestrators to kernel optimization and MLOps pipelines, indicates a broad technical curiosity and adaptability. Their academic background with an AI & ML specialization and national-level certification (GATE DA) further reinforces a dedicated interest in the field. The emphasis on production-grade systems and performance optimization suggests a pragmatic and impact-driven approach.
Soft Skills & Operational Fit
The candidate's project descriptions indicate a strong problem-solving aptitude, evidenced by optimizing LLM latency, eliminating race conditions, and resolving memory/compute bottlenecks. Their role as a 'Technical Peer Mentor' suggests leadership potential and a collaborative mindset. The focus on production-grade systems and real-time performance implies a results-oriented and detail-conscious approach.