AI Engineer with less than a year in LLM-powered systems & deep 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
An AI/ML developer with hands-on project experience building LLM-powered systems, multi-agent workflows, and deep learning models. Proficient in Python, PyTorch, and applied ML across multiple facets of AI. Currently completed a B.Tech in Computer Science (AI) and eager to contribute to production ML applications.
Muthoot Institute of Technology and Science
B.Tech · Computer Science (Artificial Intelligence)
August 1, 2022 – June 30, 2026
Ernst & Young (EY)
AI & Data Intern
January 1, 2026 – Present
India
SentiCore - Multimodal Emotion Recognition System
June 1, 2026 – Present
Co-authored a peer-reviewed paper on Spiking Neural Networks accepted and presented at IEEE ICSCST 2026 (Conference No. 70031), published in the official IEEE conference proceedings - validating the novelty of the proposed multimodal architecture. Achieved 79.29% accuracy and 79.24% weighted F1-score on 1,487 unseen samples by building a trimodal Spiking Neural Network (Spikformer) fusing text, audio, and visual features through bidirectional cross-attention — outperforming prior SNN-based baselines by 7-10%. Improved cross-modal alignment and addressed class imbalance by engineering a composite loss function combining Self-Adjusting Dice Loss, Soft-HGR correlation loss, and Cross-Entropy, resulting in more stable training across 6 emotion classes. Delivered end-to-end emotion-aware responses by integrating an LLM-conditioned empathetic response generator into the inference pipeline — the only compared system with this capability.
View ProjectLLM Wiki - Document AI Knowledge Base
June 1, 2026 – Present
Built a PDF ingestion and QA service by combining text extraction, OCR fallback, and structured chunking with an LLM query layer - enabling reliable question answering over large document collections. Ensured consistent, parseable model outputs by designing deterministic LLM prompt schemas with page-level source attribution, resulting in traceable and auditable answers. Reduced repeated ingestion overhead by implementing a document state tracker that skips already-processed files, resulting in faster pipeline runs on growing document sets. Improved retrieval relevance by chunking documents with configurable overlap and size parameters, allowing the system to be tuned for different document types and query patterns. Enabled flexible deployment by abstracting the LLM and vector store layers behind swappable interfaces, making it straightforward to swap models or backends without rewriting core logic.
View ProjectSpiking Neural Networks paper
IEEE ICSCST
January 1, 2026 – Present
Cultural Fit Analysis
The candidate's projects demonstrate a strong interest in cutting-edge AI technologies (SNNs, LLMs, multi-agent systems) and a drive for innovation, which aligns well with a dynamic AI engineering culture. The academic background and peer-reviewed publication indicate a commitment to continuous learning and contributing to the field. The diversity of projects (academic research, personal project, industry internship) shows adaptability and a broad interest in AI applications. However, the candidate's experience level is still early career, which might require mentorship and integration into a senior team.
Soft Skills & Operational Fit
The candidate demonstrates strong problem-solving skills through complex project implementations (e.g., composite loss functions, multi-agent LLM pipelines). The detailed project descriptions suggest good communication of technical concepts. The academic publication indicates a proactive and research-oriented mindset. The internship at EY suggests an ability to work in a professional environment and contribute to business-oriented solutions.