Sai Dheeraj Gummadi

Jul 2024 -

Working with FinOps teams to design and developing Cost forecasting model and Anomaly detection models for cost and revenue analysis across projects and clouds. Delivered insights into project profitability by comparing expenditures versus earnings, enabling data-driven decision-making and improved financial transparency.
Developed and implemented an automated mass email notification system that delivered weekly project cost dashboards to project owners, enabling proactive cost management by comparing current expenses against previous week and month, and highlighting realized and potential savings.
Designed and Developed an RAG based Chatbot for Public Safety Department using ChatGPT-4o-mini and AlloyDB vector database and Deployed as a RESTful API service in a Kubernetes cluster, with average response latency of 3 sec per question.
Developed an IRD Document Parser using Gemini-1.5-Flash model combined with approaches like self-consistency prompting and Deployed the parser as a RESTful API service in a Kubernetes cluster, ensuring scalability, resilience, and efficiency of the solution. This drastically reduced document extraction time from two weeks to 2 days, achieving an automation rate of over 70+%.

Nov 2023 - Jul 2024

As a Deep Learning Engineer at Brane, I Designed and Developed the pipeline for freeflow text based No Code Low Code Platform using Natural Language Processing techniques and LLMs which can build the complete solution from the given set of sentences in natural language.
Synthesised data of workflows in different domains using ChatGPT, LLama2, and Mistral for training T5/BART/GPT models for coreference resolution, paraphrasing, splitting complex sentences, span extraction.
Implemented knowledge distillation to create distilled versions of LLMs and deployed the models as RESTful APIs using Docker and Ray Serve.
Fine-Tuned GPT2 (1.5 billion parameter) for next task prediction using LORA and Bitsandbytes on multiple GPUs (Tesla V100) using model and data parallelization. A virtual assistant tailored for solution developers, aimed at enhancing their productivity.
Deployed the model with vLLM Inference Engine and achieved low-latency (0.06 sec) and high-throughput (71 rps) inferences by optimizing real-time interactions.
Prototyping RAG based approach for next task prediction using LlamaIndex for getting less hallucinated responses from LLMs.

Aug 2021 - Oct 2023

As Data Scientist at HighRadius, Designed and Developed an end-to-end pipeline for a Financial DocumentScanner product using CNN, YOLO, and OCR for extracting the required tabular regions like balance sheets, and income statements from Financial Statements and analyzing the customer’s credit risk.
Trained and Improved the 6-layered unconventional CNN using Explainable AI tools like Lime and GradCam
Achieved an overall automation rate of 80% and Reduced 40% of the time spent on manual analysis of financial documents with 100–200 pages.
Improved the latency of the Generic Parser product which is one point solution for parsing Claims/Invoices/Remittences/ POD’s using NLP transformer-based quantized LayoutLM token classification model and unsupervised algorithms with an overall accuracy of 90% and achieved automation rate of 75%.
Worked on POC for classifying the pages in input documents into Sales Invoice, Remittance, Check, and Claim for improving the accuracy of downstream parsing modules by Fine Tuning the LayoutLM Sequence Classification model for classifying the images based on both content and layout.
Developed Payment date prediction models using regression models like Random Forest / XGBoost for FORTUNE 500 CPG companies to predict invoice payment date with an average +/-3 Day accuracy of 85%.

2023 - 2024

Master's in Data Science at the International University of Applied Sciences, Berlin, Germany. I completed my Full-Time masters in distance learning format. Worked on "AI Stock Analyst: Financial Chatbot Using Large Language Models" for my Thesis under Dr.Tobias Broweleit, which talks about investment banking business, LLM's, RAG and System Design of ChatBot. Gemini 1.5 Flash and GPT-4o models were compared for stock analysis using Trustworthiness Score as evaluation metric.

2018 - 2022

B.Tech at the SRM University, AP with a major in Electronics and Communication Engineering. I Published 8 conference papers on deep learning published in IEEE, Springer, Scopus databases and Patented a pipeline for disease diagnosis under Professor Anirban Ghosh.

publications

TechRixv, 2025

G.S.Dheeraj.

TechRixv, 2025

G.S.Dheeraj.

International Conference on Intelligent Systems and Applications, 2023

G.S.Dheeraj, Anirban Ghosh.

International Conference on Innovations in Computational Intelligence and Computer Vision, 2023

G.S.Dheeraj, Anirban Ghosh

World Conference on Applied Intelligence and Computing, 2022

G.S.Dheeraj, Anirban Ghosh.

International Symposium on Digital Forensics and Security, 2022

G.S.Dheeraj, Anirban Ghosh, V.Yeswanth

International Conference on Computational Intelligence and Communication Networks, 2021

G.S.Dheeraj, V.Yeswanth, Anirban Ghosh, P.N.Kartheek, K.A.Krishna.

International Conference on Computing, Communication, and Networking Technologies, 2021

G.S.Dheeraj, Anirban Ghosh

Ripublications, International Journal of Engineering Research and Technology (Journal)

Ch.S.Nived, A.R.Kumar, G.S.Dheeraj, P.Jithendra

IRJET (Journal), 2020

G.S.Dheeraj, K.B.V.Lakshmi, K.A.Krishna

Also on Google Scholar

featured talks

Deep Learning for Disease Diagnosis at Universidad Nacional Federico Villarreal, Lima, Peru, South America. Link
CNN for Disease Diagnosis using Radiography images at IIT Kharagpur, India. Link
Residual Learning for Detecting Deepfakes at Soft Computing Research Society. Link
Compact Vision Transformer for medical images at Manipal University, Jaipur. Link
Deepfakes and Gradient Centralization for Improving Disease Diagnosis at Institute of Artificial Intelligence, Pune. Link

Machine Learning Engineer @ Motorola