Featured Work
Selected Projects
Decision systems, analytics frameworks, and agentic tools built across insurance and EV sectors.
AI Reporting Framework
Agentic + GenAI · LangChain · RAG
Standardised executive reporting using LangChain-powered agentic workflows, RAG over a VectorDB knowledge base, and reusable insight templates — reducing manual reporting overhead by ~40%.
Key outcomes
- Reduced manual report-generation effort by approximately 40%.
- Improved narrative consistency and governance across all executive dashboards.
Natural Language → Multi-Agent Analytics
Funnel Analysis + Diagnostic Agent · LangGraph
Built a LangGraph-orchestrated multi-agent system that routes plain English questions to specialised analytics agents — a Funnel Analysis agent and a Diagnostic RCA agent — powered by LLMs via the OpenAI API.
Key outcomes
- Reduced time-to-insight for ad hoc funnel and diagnostic questions.
- Enabled non-technical stakeholders to get structured funnel diagnostics without writing SQL.
Capabilities
Skills & Expertise
From raw SQL to agentic systems — tools I use to build measurable decision workflows.
Product Analytics
Data & SQL
BI & Dashboards
GenAI & Agents
Engineering
Career
Work Experience
3+ years across insurance (AXA XL) and mobility (Ola Electric, Appy High), building analytics and AI systems.
Senior Business Analyst
AXA XL
- Designed and deployed SQL-driven performance dashboards — augmented with AI-generated executive narratives — tracking profitability, risk, and growth metrics, improving pricing and underwriting accuracy by 15%.
- Built Python automation pipelines with integrated anomaly detection logic to monitor claim frequency and policy renewals, reducing manual reporting effort by 40% and enabling faster operational response.
- Ran cohort and segmentation analysis using advanced SQL to surface high-risk and high-loss segments; outputs fed directly into agentic diagnostic workflows for portfolio-level root cause analysis, improving profitability by 10%.
- Architected an end-to-end AI reporting framework using LangChain, RAG over a VectorDB, and the OpenAI API — standardising executive narratives, enforcing governance, and cutting report-generation effort by 40%.
Business Analyst
Ola Electric
- Built the Lead Funnel Management product analytics system — tracking stage-wise conversion, leakage, and daily decision cadence — achieving 15% higher lead efficiency and 10% reduction in customer acquisition cost.
- Designed the Referral Program analytics framework with end-to-end attribution tracking, funnel drop-off analysis, and experiment measurement — driving 20% growth in referral-led acquisition and 5% revenue uplift within 3 months.
- Partnered with product and growth teams to run data-driven experiments on the affiliate marketing program — SQL-based cohort analysis and A/B measurement — improving user engagement by 12% and conversion efficiency by 4%.
- Conducted VAHAN database analysis at RTO and city levels using advanced SQL to compute EV/2W market share and competitive splits, driving 9% increase in market penetration across key cities.
Get in touch
Let's build something.
Open to AI Product Analyst, Product Analytics, and GenAI-enabled BI roles. Happy to chat about interesting problems.