Akhil M.

Key Skills

Software

Don't disclose

Other

AWS SageMaker

Top Subject Matter

Healthcare compliance (FDA-reportable case identification)

FMCG & CPG Analytics

Gen AI - RAG, MCP, Fine tuning, Red Teaming

Top Data Types

Text

Image

Document

Top Task Types

Question Answering

Function Calling

Fine-tuning

Classification

Freelancer Overview

Associate Lead Data Scientist — AI Compliance Solution for a major Healthcare Provider. Brings 7+ years of professional experience across compliance, analytics, Gen AI, and structured analysis. Core strengths include AWS SageMaker. Education includes Master of Technology, BITS Pilani (2026) and Bachelor of Technology, VIT University (2017). AI-training focus includes data types such as Text, Computer Code, and Programming and labeling workflows including Question Answering, Function Calling, and Fine-tuning.

Labeling Experience

Associate Lead Data Scientist - Sigmoid Analytics

AWS SageMakerTextClassification

Led the development, performance assessment, and production deployment of machine learning and AI models for enterprise use cases. Monitored deployed models and ETL pipelines to ensure reliability, efficiency, and maintainable system behavior. Documented architectures, processes, and code to enable clear handoffs and long-term maintainability using modern ML engineering practices. • Developed and deployed LLM solutions with RAG for automated compliance classification • Monitored production model health and ETL pipeline performance • Designed and maintained scalable data/ML workflows • Produced technical documentation for systems and codebases

2025 - Present

Associate Lead Data Scientist — AI Compliance Solution for a major Healthcare Provider

Don't discloseTextQuestion Answering

Led development and deployment of an LLM using RAG to analyze client–customer communications for regulatory compliance. The system automatically identified and classified FDA-reportable cases from textual communications. Monitored production model and ETL pipeline reliability and efficiency while documenting architectures for maintainability. • Built and deployed an LLM with retrieval-augmented generation for compliance case detection • Automatically identified and classified FDA-reportable cases from communication text • Improved detection of reportable cases by 7x to reduce regulatory risk • Reduced manual compliance review costs by 50% via automated detection

2025 - Present

Software Development and Machine Learning Engineer - Vectra Automation

AWS SageMakerImageClassification

Collaborated with project stakeholders in India and the USA to define AI objectives and align prototypes with strategic requirements. Optimized algorithms to meet performance targets while satisfying regulatory constraints and operational compliance. Supervised engineering delivery through daily code reviews, feedback, and mentorship, ensuring high-quality team execution. • Built AWS Glue and PySpark pipelines to reduce processing time by 90% • Deployed production-grade agentic RAG systems for quality control and anomaly detection • Led R&D for graph neural network solutions using PyTorch and AWS SageMaker • Trained and deployed a ResNet-50 computer vision model using TensorFlow-Keras to improve classification accuracy

2021 - 2025

Software Development and Machine Learning Engineer — ResNet-50 part classification

OtherFine-tuning

Trained and deployed a computer vision model to perform part classification for manufacturing workflows. The model produced labeled class outputs for parts and was validated for high classification accuracy. The deployment reduced CAD engineers’ workload by automating part identification tasks. • Trained and deployed a ResNet-50-based computer vision model for part classification • Achieved 98% accuracy in part classification • Reduced CAD engineers’ workload through automated classification • Delivered production-ready deployment using TensorFlow-Keras workflows

2021 - 2025

Software Development and Machine Learning Engineer — Deployed production LLMs with agentic RAG

OtherFunction Calling

Deployed production-grade agentic RAG systems that performed automated quality control and anomaly detection using industrial data in automotive manufacturing. The work involved designing AI workflows to extract relevant information and trigger appropriate model behaviors for downstream actions. Supervised team productivity via daily code reviews, feedback, and mentorship during ongoing model development. • Deployed agentic RAG pipelines for quality control and anomaly detection • Automated monitoring tasks to improve operational efficiency by 20% • Coordinated with project managers to align AI objectives with strategic goals • Provided technical leadership through code reviews, mentoring, and feedback

2021 - 2025

Education

Work History