Looking for a job

Nikunj Brahmbhatt

@nikunjbrahmbhatt

Enterprise AI Value Strategy Consultant at Accenture; fine-tuned an LLM that cut bank-comparables inference costs by approximately 1,100x.

United States

Message

At Accenture, I advise C-suite leaders on AI and data strategy, assess organizational maturity, and shape transformation roadmaps. I translate AI products and services into adoption plans, business cases, and ways to track value.

In an open-source project, I fine-tuned Llama 3.1 8B for bank comparable-companies analysis, with zero numerical hallucinations and approximately 1,100x lower inference costs than the benchmark cited in my project. I also built BankerTrainer, a local voice-enabled AI training simulator for retail bank associates.

Experience

Work history, roles, and key accomplishments

Current

Fine-Tuned LLM for Bank Comparables Analysis

Current

Fireworks AI

May 2026 - Present (5 months)

Fine-tuned Llama 3.1 8B (SFT + LoRA) to automate a bank comparable-companies workflow, achieving 100% source-citation discipline and zero numerical hallucinations. Shipped end-to-end in 5 days on $35 of compute.

Fine Tuning SFT LoRAs LLaMA 3.1 Evaluation Frameworks RAG Python GitHub

Current

BankerTrainer - Voice-Enabled AI Training Simulator

Current

Independent Build

Jan 2026 - Present (9 months)

Engineered a 100% local, voice-enabled simulator for retail bank associates using Llama 3.1 8B, Whisper, Piper, and Streamlit. Layered deterministic guardrails and post-parse output repair, attaining 18 of 18 correct outcomes in stress testing.

LLaMA 3.1 Ollama Whisper PIPER Streamlit Voice AI GuardRails Python

Current

Enterprise AI Value Strategy Consultant

Current

Accenture

Jul 2024 - Present (2 years 3 months)

Serves as trusted advisor to C-suite leaders, shaping AI and data strategy and transformation roadmaps. Owns client engagements end-to-end, delivering executive readouts and strategic analyses.

AI Strategy Data Strategy GenAI LLM RAG AI Agents Business Case Development Executive Stakeholder Management Azure AI AWS Databricks Snowflake Responsible AI

LLM Evaluation for Financial Modeling

Micro1 AI

Jan 2026 - Feb 2026 (1 month)

Stress-tested Claude, ChatGPT, and Gemini across 40+ DCF, LBO, and merger-model prompts against a 15-point rubric, surfacing a 22% accuracy gap in AI-generated sensitivity tables. Informed the lab's finance-model training pipeline.

LLM Evaluation Claude ChatGpt Gemini Financial Modeling DCF LBO Merger Modeling Prompt Engineering

Digital Transformation Analyst

Agile Infoways Pvt. Ltd.

Jun 2019 - Jul 2022 (3 years 1 month)

Partnered with enterprise clients to assess processes and technology maturity, converting business objectives into scoped cloud, data, and automation initiatives. Managed strategic accounts end-to-end and produced performance dashboards and revenue forecasts.