🎓 MSc Artificial Intelligence (Distinction) — Sheffield Hallam University, UK 💼 4+ years of end-to-end ML engineering experience across NLP, computer vision, and cloud-deployed inference 🔍 Currently building Generative AI and Agentic AI systems at Synechron — LLM orchestration, MCP-based tool integration, and enterprise-scale AI for financial services clients 📍 Based in Bengaluru, India | 🗺️ Open to senior AI/Data Science roles across the UK, Europe, UAE, Canada, and New Zealand — remote or relocation

Previously at Quantiphi, where I delivered NLP automation pipelines that improved claim adjudication efficiency by 80%, and at Winjit Technologies, where I built computer vision systems contributing to ~50% revenue growth for enterprise clients including Mahindra & Mahindra.

Personal automation pipeline for a UK Skilled Worker sponsorship search — reads public government data and drafts content; every risky action (applying, sending, messaging) stays a manual human click by design

• MCP server (src/mcp_server/) exposing the pipeline as four tools — check_sponsor, check_salary_threshold, track_application, list_applications — so an MCP client like Claude Desktop can query and update the search conversationally, verified end-to-end against a real client
• Thin-wrapper MCP architecture: zero business logic in the tool layer, connection-per-call SQLite access, and project-root-anchored paths — built to survive an MCP client spawning the server from an arbitrary working directory
• UK sponsor-register ingestion with name normalization (trading names, LTD/LIMITED/LLP/PLC suffix variants) against the real gov.uk register, plus SOC-code salary-threshold checks against actual government thresholds
• Documented spec-and-review discipline: intent contract, I/O edge-case matrix, and an adversarial review triage log per feature; 122 passing tests

Full-stack hiring-evaluation platform that assesses candidates on real-world AI-assisted coding competency, not just algorithmic recall

• Candidates solve challenges in an isolated Docker + browser-VS-Code environment with Gemini CLI access; full workspace and AI-session transcript are captured on submit
• LLM-driven 8-dimension scoring (problem decomposition, debugging with AI, architecture decisions, and more), with hire thresholds enforced in Python, not by the LLM's own claim
• Employer dashboard with radar-chart candidate comparison, human override/flag workflow, and an append-only audit trail
• 11-table SQLite schema, 199 passing tests (fully mocked LLM/Docker), and a benchmarking suite run against the real Gemini API and real Flask routes — with documented, file-referenced production bottlenecks (N+1 query, connection pooling, single-worker WSGI) found via that benchmarking, not guessed

GenAI-powered conversational analytics using LLM-driven SQL generation and multi-agent workflows

• Natural-language querying over retail datasets via schema-aware SQL generation
• Multi-agent reasoning with conversation memory for contextual queries
• Designed for 100GB+ dataset scale

End-to-end Retrieval-Augmented Generation system for document Q&A

• LangChain + OpenAI Embeddings + FAISS vector database
• Document ingestion, chunking, vector indexing, and grounded LLM responses with citations
• FastAPI backend for scalable document question-answering

AI agent that auto-analyzes CSV datasets using Claude AI + Machine Learning

• Plain-English queries → automated ML analysis pipeline
• Intelligent tool selection for classification, regression, and clustering

NLP-based resume parsing with SpaCy NER and rule-based extraction

• Parses PDF, DOCX, TXT resumes into structured JSON/CSV
• Section-aware extraction for skills, experience, and education
• FastAPI batch-processing API

LLM & Agentic AI

Languages & ML Frameworks

Cloud & Deployment

Data & Tools

• MSc Artificial Intelligence (Distinction) — Sheffield Hallam University, UK (2022 – 2023)
• Post-Graduation Diploma, Machine Learning and AI — IIIT Bangalore (2020 – 2021)
• B.E., Information Technology — Sinhagad Technical Education Society's COE (2013 – 2018)

• ☁️ AWS Certified Developer – Associate
• 🤖 IBM Generative AI for Data Scientists Specialization — Coursera
• 📈 Structuring Machine Learning Projects
• 🐍 HackerRank Certified Python
• ⛓️ Introduction to Blockchain
• 🧠 Machine Learning and AI

💬 Open to senior AI/Data Science roles across the UK, Europe, UAE, Canada, and New Zealand — remote or relocation.
📩 Reach me at Kunal.gaikwad06@outlook.com