Sidharth Ramachandran profile photo 1 on Tal

Sidharth Ramachandran

Data Scientist at Air India | Artificial Intelligence | Machine Learning | Deep Learning | NLP | Computer Vision | Data Engineering | Working on RAG frameworks, large language model tuning, and generative AI solutions

Kochi

5+ years of work experience

Air India Limited
Aug 20232 yrs 11 mosPresent

Kochi

Building production-grade AI systems at scale for one of India's largest full-service carrier, serving millions of customers. Work spans GenAI and RAG, deep learning, agentic pipelines, safety-critical audio analytics, and ML systems trained on millions of customer records. Actively building POCs on emerging AI technologies including agentic frameworks, MCP, and multimodal LLMs. Enterprise RAG Question-Answering System: Architected and deployed an enterprise-grade RAG-based QA platform covering three domains: customer experience queries, internal flight delay information, and social media intelligence from LinkedIn and X. Built semantic search using vector embeddings with LLM-powered response generation via LangChain, enabling real-time resolution at scale. Social Media Analytics at Scale: Built a large-scale analytics platform using LLMs and GenAI techniques to monitor public sentiment, track complaints, and surface emerging issues across LinkedIn and X. Deployed to support the Vistara merger, tracking brand perception and customer sentiment during a high-stakes transition. Post-Call Analytics Platform: Built an LLM-based pipeline processing tens of thousands of daily call transcripts. Applied unsupervised clustering to surface recurring issues and extract actionable operational insights. Deployed via batch inference for cost-efficient processing at scale, reducing operational costs by 50%. CVR Safety Analysis System: Built a safety-critical audio analytics pipeline for Cockpit Voice Recorder data, scaled via Spark-based parallel processing. Extracted safety insights using LLMs for aviation compliance and audit workflows. Developed a CNN-based model to detect vocal disfluencies and cognitive load markers in flight recordings for flight safety analysis. Customer Travel Propensity Model: Built a propensity model trained on millions of customer records, deployed as a batch scoring pipeline integrated with marketing systems.
VensureHCM
Nov 202210 mosAug 2023

Kerala, India

(Rebadged from Gritstone Technologies) Designed and deployed production ML systems across recommendation, NLP, and generative AI domains. Focused on translating business problems into scalable microservice-based ML architectures with measurable product impact. Contribution Recommendation System : Built an ML-powered recommendation system to optimize user contribution decisions. Engineered features and applied constraint-based optimization to align recommendations with real-world business rules. Delivered measurable improvements in contribution rates and benefit capture. Deployed as a Flask microservice. NLP Data Pipeline :Designed and implemented NLP pipelines to transform raw text data into structured, ML-ready formats. Covered end-to-end preprocessing, feature extraction, and integration with downstream model training workflows. Generative AI Integration : Built generative text capabilities including summarisation, text generation, translation, and question answering using LangChain and LLMs. Integrated into product workflows to automate and enhance text-heavy processes.
Gritstone Technologies
Jul 20212 yrs 2 mosAug 2023

Kozhikode

Owned the full ML lifecycle across production systems, from scoping and EDA through deep learning and NLP model development, evaluation, and cloud deployment. Built and shipped real-world AI pipelines with measurable business impact. NLI Consistency Engine : Fine-tuned a RoBERTa cross-encoder via transfer learning for 3-way classification on a curated hypothesis-premise pair dataset, achieving strong benchmark accuracy. Designed to cross-validate candidate answers against questions for logical and factual grounding inside an interview platform. Deployed as a low-latency FastAPI inference service for real-time response validation. Document Auto-Population Pipeline :Built an end-to-end NLP and OCR pipeline using spaCy and EasyOCR to extract and auto-populate fields from scanned documents. Cut document processing time by over 95%, enabling near real-time throughput.
European Water Technologies India
Jan 20212 mosFeb 2021

Kochi, Kerala, India

Education

Cochin University of Science and Technology

Master's degree, Marine Geophysics

Cochin University of Science and Technology

20172 yrs2019

Bachelor of Science - BS, Physics, Astrophysics

Nirmalagiri College (Autonomous), Kuthuparamba

20133 yrs2016

Join Sidharth and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Sidharth Ramachandran on Tal

Headline

Data Scientist at Air India | Artificial Intelligence | Machine Learning | Deep Learning | NLP | Computer Vision | Data Engineering | Working on RAG frameworks, large language model tuning, and generative AI solutions

About

ML Engineer and Data Scientist passionate about leveraging data science and machine learning to drive impactful solutions. Specializing in deep learning with a keen interest in exploring adversarial attacks for enhancing defensive neural networks. Additionally, fascinated by the intersection of AI and neuroscience, and how insights from neuroscience can inform and enrich AI algorithms. Prior experience in product engineering provided a solid foundation for tackling complex problems. Always eager to collaborate and contribute to cutting-edge projects in the field.

Location

Kochi

Work experience

Data Scientist at Air India Limited, Aug 2023 - Present · 2 yrs 11 mos, Kochi, Building production-grade AI systems at scale for one of India's largest full-service carrier, serving millions of customers. Work spans GenAI and RAG, deep learning, agentic pipelines, safety-critical audio analytics, and ML systems trained on millions of customer records. Actively building POCs on emerging AI technologies including agentic frameworks, MCP, and multimodal LLMs. Enterprise RAG Question-Answering System: Architected and deployed an enterprise-grade RAG-based QA platform covering three domains: customer experience queries, internal flight delay information, and social media intelligence from LinkedIn and X. Built semantic search using vector embeddings with LLM-powered response generation via LangChain, enabling real-time resolution at scale. Social Media Analytics at Scale: Built a large-scale analytics platform using LLMs and GenAI techniques to monitor public sentiment, track complaints, and surface emerging issues across LinkedIn and X. Deployed to support the Vistara merger, tracking brand perception and customer sentiment during a high-stakes transition. Post-Call Analytics Platform: Built an LLM-based pipeline processing tens of thousands of daily call transcripts. Applied unsupervised clustering to surface recurring issues and extract actionable operational insights. Deployed via batch inference for cost-efficient processing at scale, reducing operational costs by 50%. CVR Safety Analysis System: Built a safety-critical audio analytics pipeline for Cockpit Voice Recorder data, scaled via Spark-based parallel processing. Extracted safety insights using LLMs for aviation compliance and audit workflows. Developed a CNN-based model to detect vocal disfluencies and cognitive load markers in flight recordings for flight safety analysis. Customer Travel Propensity Model: Built a propensity model trained on millions of customer records, deployed as a batch scoring pipeline integrated with marketing systems.

Artificial Intelligence / Machine Learning Product Engineer at VensureHCM, Nov 2022 - Aug 2023 · 10 mos, Kerala, India, (Rebadged from Gritstone Technologies) Designed and deployed production ML systems across recommendation, NLP, and generative AI domains. Focused on translating business problems into scalable microservice-based ML architectures with measurable product impact. Contribution Recommendation System : Built an ML-powered recommendation system to optimize user contribution decisions. Engineered features and applied constraint-based optimization to align recommendations with real-world business rules. Delivered measurable improvements in contribution rates and benefit capture. Deployed as a Flask microservice. NLP Data Pipeline :Designed and implemented NLP pipelines to transform raw text data into structured, ML-ready formats. Covered end-to-end preprocessing, feature extraction, and integration with downstream model training workflows. Generative AI Integration : Built generative text capabilities including summarisation, text generation, translation, and question answering using LangChain and LLMs. Integrated into product workflows to automate and enhance text-heavy processes.

Product Engineer ML/AI at Gritstone Technologies, Jul 2021 - Aug 2023 · 2 yrs 2 mos, Kozhikode, Owned the full ML lifecycle across production systems, from scoping and EDA through deep learning and NLP model development, evaluation, and cloud deployment. Built and shipped real-world AI pipelines with measurable business impact. NLI Consistency Engine : Fine-tuned a RoBERTa cross-encoder via transfer learning for 3-way classification on a curated hypothesis-premise pair dataset, achieving strong benchmark accuracy. Designed to cross-validate candidate answers against questions for logical and factual grounding inside an interview platform. Deployed as a low-latency FastAPI inference service for real-time response validation. Document Auto-Population Pipeline :Built an end-to-end NLP and OCR pipeline using spaCy and EasyOCR to extract and auto-populate fields from scanned documents. Cut document processing time by over 95%, enabling near real-time throughput.

Data Scientist at European Water Technologies India, Jan 2021 - Feb 2021 · 2 mos, Kochi, Kerala, India

Education

Master's degree, Marine Geophysics, Cochin University of Science and Technology

Bachelor of Science - BS, Physics, Astrophysics, Nirmalagiri College (Autonomous), Kuthuparamba