Pranjal Chaubey

As a Business Analyst in AI Transformation, I am an ambitious and independent professional, deeply committed to accountability and the pursuit of mastery in my work. My pragmatic approach, balanced with a strong sense of empathy, drives me to deliver impactful solutions and continuously refine processes for optimal outcomes.

Experience

Business Analyst, AI Transformation

Tata 1mg · Jul 2025 – Present

Not yet confirmed

• Promoted from AI Engineering Intern (Jul 2025 – Mar 2026) in Apr 2026.
• Team charter: enter a department, map the real process, rebuild it around AI systems.
• Optimus (enterprise AI copilot).
• Cut HR query load ~30% across 20 business partners (~6 FTE released) by building an enterprise agentic RAG copilot for 6,000+ employees: ReAct with RAPTOR and CRAG retrieval on Qdrant, FastAPI, Redis and OpenAI, shipped inside the chat tool staff already used rather than as a separate app.
• Webassure (catalog automation).
• Raised catalog compliance from ~70% to >98% and reduced ops headcount 16 to 4 (INR 5L+/month) by building LLM structured extraction of 64 fields from low-quality product images, with hybrid dense-embedding and fuzzy entity resolution across 800K+ SKUs.
• Webassure, extended.
• Prevented INR 1.5Cr+ in GST leakage across 160K misclassified SKUs and scaled onboarding to 10K+ SKUs/month by extending that pipeline to rate prediction, Rx compliance and SEO.
• Featured publicly by the Tata 1mg CTO.
• MER Audit (multimodal QC pipeline).
• Scaled insurance call auditing 300 to 1,500+ calls/day at ~98% accuracy, cutting turnaround 67% and review headcount 27 to 11, by building an orchestrated pipeline (ElevenLabs transcription, Gemini discrepancy detection across 186 questions) that delivers reviewer-ready analysis ~15 minutes after a call ends, escalating only exceptions to a human and auto-submitting clean cases.
• PII-compliant end to end.
• Retail Audio Analytics.
• Added INR 3L/month incremental revenue per store across 15 stores by building multi-signal SKU retrieval (BM25 + dense embeddings + phonetic + fuzzy, fused via Reciprocal Rank Fusion, LLM-reranked) surfacing out-of-stock and cross-sell signals.
• AI governance.
• Own enterprise AI governance (Langfuse observability, MLflow tracking, Skyflow PII) and technical evaluation of every AI vendor the company adopts; internal technical reference for teams introducing AI.

Data Scientist

Tata 1mg · Jan 2025 – Present

Not yet confirmed

Data Engineer Intern

MathCo · Jan 2025 – Jul 2025

Not yet confirmed

• Tech Stack: Azure · Snowflake · Python · SQL · Jinja · Git · Data Modeling · System Design
• • Designed and implemented a Single Source of Truth (SSOT) and enterprise data mart using Snowflake and Azure Data Factory, unifying client data.
• • Developed scalable ETL workflows with data quality checks and automated monitoring, enhancing data reliability.
• • Created reusable SQL templates with Jinja and Git, streamlining transformation logic and environment management.

M

Data Engineering Intern

MATHCO · Jan 2025 – Jun 2025

Not yet confirmed

• Built ETL pipelines feeding enterprise analytical data marts on Snowflake (Azure Data Factory, Snowflake SQL, Python), with parameterized Jinja2 transformations, CI/CD, and automated data-quality and schema-drift checks.

What drives their work

The Architect

You design and construct robust systems that transform complex, manual processes into efficient, AI-native operations.

Pranjal's superpower

You excel at transforming complex, manual business processes into highly efficient, AI-driven systems with meticulous attention to accuracy and scalability.

How Pranjal works

You are best suited for roles in dynamic teams focused on building and deploying AI/data solutions that drive significant business transformation.