[Your Name]

Architected & Deployed Multi-Agent GenAI Chatbot: Led the development of a multi-agent analytics chatbot (RAG architecture), reducing manual data analysis time for clients by an estimated 90%.

Engineered an Active ReAct framework with a Qdrant vector store and an RLHF feedback loop to ensure consistent, high-quality, and continuously improving responses.

Launched AI-Powered Call Analytics Product: Spearheaded the launch of a new Call Analytics product, reducing manual audit time for a key insurance client from 3 hours to just 30 minutes per call.

Architected a scalable backend pipeline with a self-hosted Whisper Large-v2 model on GPU infrastructure to deliver high-accuracy, scalable quality assurance.

Delivered AI-Powered Summarization & Search: Shipped a new AI feature providing topic-level summaries from thousands of comments, saving users over 3 hours of manual analysis per report.

Implemented semantic search to power the feature and enhance data exploration.

Automated Critical Business Operations: Deployed N8N automation workflows that slashed operational task times by up to 90%; reducing survey generation from 2 hours to 15 mins, data generation from 1 hour to 5 mins, and CSS creation from 5 hours to 1 hour.

Implemented End-to-End CI/CD Pipelines: Automated the ML deployment lifecycle by building CI/CD pipelines that triggered deployments to UAT and Production environments, eliminating manual deployment errors and accelerating release cycles.

Advanced Model Customization & Cost Management: Initiated the fine-tuning of Llama 3.1 (QLoRA) on proprietary data and developed a custom framework to monitor LLM costs by client, user, and agent, providing critical business intelligence via a Streamlit dashboard.

Pioneered Company's GenAI Strategy: Drove the initial adoption of Generative AI by extensively researching prompt engineering and building multiple functional RAG prototypes (POCs).

Proved the technology's business viability, which led to the funding and development of the flagship Analytics Chatbot.

Led Strategic AI Initiatives: Collaborated directly with business leadership to define the roadmap for new Predictive Analytics features, directly influencing product direction and securing client buy-in.

Launched Key Driver Analysis Product: Owned the full lifecycle of the 'Key Driver Analysis' feature, from statistical modeling to backend development, creating a new analytics module that identified the primary drivers of customer NPS.

Developed Novel NLP Features: Launched a 'Text Analytics Actionables' feature using Sentence Transformers and T5 models, automatically extracting and clustering actionable business insights from unstructured customer feedback.

Scaled Team & Delivery: Hired and mentored a data scientist, successfully leading the team to deliver numerous custom dashboards and features for key enterprise clients.

Scaled ML Infrastructure with MLOps: Containerized all ML applications using Docker and orchestrated their deployment on Kubernetes, improving system reliability and deployment velocity.

Reduced Model Inference Costs by >60%: Optimized core NLP models by migrating from BERT to DistilBERT with int4 quantization, significantly lowering operational expenses without sacrificing accuracy.

Contributed to Open Source: Submitted a validated bug fix to the simpletransformers library for a quantized NER model prediction error, which was merged into the main branch.

Invented Novel Sentiment Algorithm: Developed a proprietary keyword-level sentiment algorithm that improved accuracy on complex sentences by over 30% compared to the baseline sentence-level model.

Upgraded Core Analytics Engine: Migrated the company’s foundational text analytics from a legacy TF-IDF system to a BERT-based model, dramatically improving performance and enabling new product capabilities.

Built Foundational Data Pipelines: Deployed services on AWS and implemented a Redis-based message queue to handle asynchronous processing, improving system throughput and scalability.