Headline

Senior Data Scientist at HCLTech | I build scalable AI solutions that drive business growth | Proficient in deep learning & cloud-based data science pipelines using AWS/Azure/GCP

About

Experienced AI/ML Engineer and Architect with over 11 years in the industry, delivering scalable products and solutions for multinational clients across the US, Europe, Middle East, and APAC. Specializing in application-driven AI systems, LLM architectures, RAG solutions, and end-to-end ML engineering with proven business impact. Expert in designing and deploying production-grade AI offerings (LLMs, generative models, MLOps pipelines) using Python, PySpark, SQL, React, and cloud platforms (AWS, Azure, Cloudera, GCP). Extensive experience with AI-driven medical coding analysis projects, clinical data systems, and cross-functional data engineering for healthcare clients, including LAD documentation, PDX medical coding, and compliance with HIPAA, GDPR, and ISO standards. Proficient in remote, hybrid, and multicultural distributed teams using Jira, Slack, GitHub, and CI/CD tooling. Key Achievements: - Architected LLM workflows and RAG systems for enterprise clients in finance, healthcare, and technology sectors across continents - Built medical coding analytics and documentation solutions for global healthcare providers and insurance clients - Delivered remote-ready platforms and APIs, including medical coding dashboards and compliance reporting, adopted in EMEA and Asian regions - Led global teams and coordinated with stakeholders in multiple time zones to achieve project goals ahead of deadlines - Provided mentorship, peer training, and contributed to open-source ML/AI projects for global communities Open to both remote and onsite opportunities in Europe, China, Middle East, Finland, and worldwide. Willing to relocate and available for visa sponsorship as needed. Currently enhancing my skills in Finnish and Mandarin for international collaboration.

Location

Kolkata

Work experience

Technical Architect — Agentic AI for Technical Documentation | Siemens DISW (CPSM) at HCLTech, Nov 2025 - Present · 7 mos, India, Built an end-to-end agentic AI system for intelligent technical writing assistance, targeting DITA-based authoring platforms (IXIA/CCMS). The system orchestrates multi-source context retrieval via MCP connectors across Jira, Polarion, Amplitude (customer feedback), and S3 artifacts — scoped per author to surface relevant requirements, defects, and user signals at authoring time. Designed a hybrid rule generation pipeline combining Schematron-based structural validation with LLM-generated semantic rules to detect errors and quality issues in DITA XML documents at scale. The agentic workflow coordinates retrieval, rule evaluation, and AI-assisted drafting to guide technical writers through the creation of complex engineering topics — reducing dependency on manual cross-referencing across fragmented toolchains. Key contributions: • Architected the multi-connector MCP retrieval layer across structured (Jira/Polarion) and unstructured (S3/Amplitude) sources • Designed the rule synthesis pipeline (Schematron + LLM hybrid) for DITA document quality enforcement • Built the agentic orchestration layer coordinating retrieval, validation, and generation steps • Integrated with IXIA CCMS as the authoring target platform Stack: Agentic AI · MCP · DITA/XML · Schematron · LLMs · Jira · Polarion · Amplitude · S3 · IXIA CCMS · Siemens DISW

Senior Data Scientist | HCL Technologies@AbbVie at HCLTech, Jul 2024 - Oct 2025 · 1 yr 4 mos, Lead a team of 3 analysts across two production AI systems in medical coding and LLM operations. Owned client-facing delivery including bi-weekly stakeholder reviews with AbbVie's Director of Data Science and clinical operations leadership. Architected a production medical coding LLM system on fine-tuned PHI-4 with LoRA/QLoRA, processing 10,000+ medical charts daily at 95% accuracy — reducing manual review time by 70% and accelerating claims processing cycles by ~20%. Built the full MLOps stack on Databricks with Delta Lake, MLflow, and Grafana observability, alongside a hybrid RAG system (dense + BM25) over 100K+ medical documents achieving 92% retrieval accuracy. Separately designed a scalable LLM-Ops log processing pipeline on Databricks handling 1M+ daily log entries, cutting alert fatigue by 70% and MTTR by 50%. Key contributions: • Fine-tuned PHI-4 with 4-bit quantization — 65% model size reduction with no inference quality loss • Hybrid RAG (Pinecone) with semantic caching — 60% latency reduction across 100K+ documents • End-to-end MLOps: Delta Lake, MLflow (50+ experiments), Hyperopt, Grafana production dashboards • LLM-powered intelligent alerting reducing MTTR by 50% across 1M+ daily log events • Led team of 3 through full model development, evaluation, and deployment lifecycle Stack: PHI-4 · LoRA/QLoRA · RAG · Pinecone · Databricks · Delta Lake · MLflow · Grafana · GPT-4 · Python · HCL Technologies

AI/ML Computational Science Specialist | RAG, Knowledge Graphs, Predictive Modeling, Snowflake at Accenture, Feb 2024 - Jun 2024 · 5 mos, India, • Led predictive maintenance and lead time detection projects for wind turbine failures using advanced sensor data analysis, Power BI dashboards, and time series ML models. Developed clustering and anomaly detection workflows in Azure ML and Databricks/PySpark, delivering proactive fault prediction systems for global energy clients. • Designed and implemented sophisticated multi-agent RAG framework for enterprise knowledge management, improving query accuracy and operational efficiency. Built AI-driven analysis solutions with advanced retrieval techniques, semantic search, and intelligent query-ranking systems. • Developed production-grade ML pipelines and MLOps frameworks using Azure ML, Databricks, and Power Platform, including model versioning, A/B testing, automated deployment, and comprehensive monitoring for enterprise applications. • Implemented strategic ML solutions for remote operations optimization, collaborating with cross-functional teams across multiple time zones. Automated data pipelines using Snowflake and Azure DevOps, significantly reducing data processing time and improving operational workflows.

Data Science Consultant | Workforce Planning | Theme Analysis at Accenture, May 2021 - Feb 2024 · 2 yrs 10 mos, Bengaluru, Karnataka, India, As a Data Science Consultant in workforce planning, I facilitated clients improve staffing strategy. By analyzing workforce data and developing predictive models in Python, I identified trends in employee turnover and projected future needs. My insights, presented through clear visualizations, led to significant improvement in workforce planning accuracy, resulting in better resource allocation and reduced hiring costs.

Senior Machine Learning Engineer | IoT Solutions, Predictive Analytics, Demand Forecasting at Quantiphi, Nov 2020 - May 2021 · 7 mos, Bengaluru, Karnataka, India, I delivered an IoT system with predictive analytics tools that addressed three key inefficiencies in inventory management workflows. This system enhanced object recognition precision by 20% through bounding box normalization, ensuring scalability with GCP custom jobs. Additionally, I developed a demand forecasting model using Prophet and DTW, improving the accuracy of inventory predictions.

Analytics Engineer (Data Science) | Fraud Detection, Auto-ML, Time Series Analysis at Impetus, Oct 2019 - Nov 2020 · 1 yr 2 mos, Bangalore, I improved fraud detection accuracy by 20% using Random Cut Forest and Prophet-based anomaly detection models. Additionally, I shortened prototyping time by 50% by implementing auto-ML workflows for pattern detection.

Masters Student at IIIT Hyderabad, Jul 2017 - Jul 2019 · 2 yrs 1 mo, Hyderabad, India

Education

Master's degree, Information Technology, International Institute of Information Technology Hyderabad (IIITH)

Bachelor’s Degree, Electronics and Communications Engineering, haldia institute of technology