Boeing

Pilot Training using Gen AI  (Sep 2023 to Present)

1.	Used RAG framework to develop learning solutions for pilots through feedback based learning 
2.	Helped and taught the team in adoption of GEN ai for any problem statement
3.	Used Open AI GPT -4, Hugging face Models like Zephyr, mistral, Llama2 to develop models 
4.	Used Vector DB’s like Fiass, Weviate and Simple vector DB of llamaindex
5.	Finetuned Existing models using Lora  and QLora PEFT techniques 
6.	Experimented the model with custom embeddings, Fusion of Opensource and closed source models using langchain and llamaindex frameworks 
7.	Developed POC’s for a complete end to end generative AI aided training solutions

Data Scientist

AB InBev

1) Create solutions using predictive modelling and inferential statistics
2) Creating solutions involving mathematics and statistics at the core of all the solutions that are developed with tensorflow, python and its libraries, 
3) I worked with Large language models and created a POC that would close the gap between entry level role to Vice president 
4) Used GPT -4 to create a project from scratch and automate multiple sceanrios
5) Taught Large language models use cases and prompt engineering to my team and got recognized for it
6) Implement them using standard OOPs with robustness and scalability as the goal
7) Working on fine-tuning and stabilizing the existing model for MMM
8) Devoloping solutions to solve market mix moulding problems
9)Predicting investments in digital vehicles that gives maximun return of investments 
10)Check for anomalies in data with respect to pricing, sales 
11)Analyze the sales in multiple countires and recommend best investments to key stake holders
12)Working with timeseries data of several places , Combining them to derive meaninful insights

Tata Consultancy Services

•	Worked with Open-CV library for morphological operations on road
•	Worked in Computer vision related tasks like image segmentation of objects in an image
•	Worked on Object Detection, Object localization using CNN and Yolo-V3 models, 
        Worked in lane detection team , used Sci-kit learn, Tensorflow, keras , DBSCAN , Unsupervised 
        Algorithms.
•	Worked in lidar object detection team, used Numpy, pandas, sea-born, matplotlib to understand, process and produce meaningful data for training.
•	Exposure to RCNN, SDD Models 
•	Involved in trained models from scratch for custom signs detection
•	Used Data Augmentation Techniques while re-training 
•	Worked in a cross-functional team, where every step has dependency on others
•	Working as sensor fusion engineer where we associate and unify all the sensor outputs and fed them to reinforcement learning agents
•     Used Multithreading/Multiprocessing to solve critical asyncronouse tasks
•     Used Azure scalable VM’s , storage blobs, created pipelines , unified them to automate the labelling procedures.

•	Used Natural language processing libraries like NLTK to automate tasks
•	Used RNN and LSTM for synchronizing the validation tasks, Have exposure of SeqtoSeq models like hugging face, Glove, transformers.
•	Consistently worked on changing requirements, Created resilient workflow, developed the code with standard OOPs concepts and tested it accordingly and also deployed into multiple systems in C# and python.
•	Understood the workflow of test procedure validation and automated the pipeline up to 75 % 
•	Used Pandas and Numpy libraries to do data processing.
•	Developed scripts to automate mundane tasks in excel word with standard OOPs concepts either in C++ and Python

Developer (NLP - Python, C#)

Talentio Solutions India Private Limited

Trained young students in C,  Data structures and in-depth Competitive programming courses specially focused for campus placements

Technical Trainer

Intern Technical trainee

I am an outoftheboxthinker ;) 

My experience is in solving people's problems using AI and Non AI methods

During this process i earned strong experience in Python with OOPS, Flask , MongodB and Natural language processing techniques (With and without Deep learning)

I also dealt with images i.e computer vision tasks which involved using pre-trained/Custom trained models

When dealing with data related problems i have used pandas, numpy, scikit learn, tensorflow, keras and visualization libraries like seaborn and matplotlib.

I am a full stack data scientist and a very passionate teacher

Taplio

Buchhibabu Rachakonda's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Taplio

.css-1jgreyo{background-image:linear-gradient(284deg, rgba(39,43,72,1) 0%, rgba(33,47,183,1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Buchhibabu Rachakonda's Linkedin Analytics

Want to drive more opportunities from LinkedIn?

Buchhibabu Rachakonda's Linkedin Analytics