**Kaggle achievements**:
- World’s first Quadruple Grandmaster
- Top 5 in all categories
- Kaggle profile: http://www.kaggle.com/abhishek

Get my book: Approaching (Almost) Any Machine Learning Problem for FREE: https://github.com/abhishekkrthakur/approachingalmost

I post a lot of cool stuff from time to time on my YouTube channel. Subscribe me on YouTube: https://www.youtube.com/AbhishekThakurAbhi

I hold a Master's Degree in Computer Science from the University of Bonn. In my free time, I like to take part in machine learning competitions and have taken part in over 200 competitions. My research interests include automatic machine learning, deep learning, hyperparameter optimization, computer vision, image analysis and retrieval and pattern recognition.

Abhishek Thakur

SalesMeetings.ai

➢ Developed scalable production pipelines leveraging Gen AI, Deep Learning (FastPitch, HiFiGAN), and AWS stack (EC2, ECS, Lambda, Redis) to support 25+ customers and scale output from 100 to 10,000+ videos per day, and enable a key client to deliver 3,000+ personalised videos daily, reducing content generation and enhancing end-user engagement.
➢ Built and deployed in-house speech synthesis models, delivering human-like AI voices for marketing videos and reducing
external service costs by 90% for a single customer.
➢ Automated and optimised media processing workflows using serverless AWS architecture to meet increasing business
demands and enable rapid client onboarding.

Software Engineer (Machine Learning)

Pirimid Fintech

➢ Led design and implementation of ML solutions, onboarding 3 new enterprise clients and driving revenue growth by 10%.
➢ Designed "Ignosis" early warning system, improving lender default prediction accuracy by 10% and reducing false positives by 5% using deep learning on structured and unstructured data.
➢ Developed NLP systems leveraging Cosine/Jaccard similarity to analyse 10-K/10-Q SEC filings, enabling data-driven investment for hedge funds.
➢ Built OCR and NLP pipelines on AWS for stock research report parsing, enhancing portfolio analytics and decision-making.

AI/Machine learning engineer

• Conducted research on AI applications in Fintech and the Indian market, by analyzing the company pipeline and assisting the team in implementing 5 such papers ideas.
• Created a cryptocurrency price prediction model using time-series analysis and machine learning algorithms which decreased the prediction error by 60% over a baseline model.
• Created an ML model to predict the trader's next action using the previous history. The resultant model predicted the next move of traders with a 65% F1 score.

Machine Learning/AI intern

Villa-Tech, Inc.

• Leveraged Deep Learning models, Gstreamer, and OpenCV in Python to upgrade computer vision applications, introducing real-time detection features like facial and object recognition, culminating in a 20% performance improvement.
• Engineered data pipelines and model training workflows using TensorFlow and Keras for streamlined ML development processes that amplified model training speed by 15% and improved model accuracy by 5%.
• Developed the model evaluation pipelines which could test the detection models in real-time thus reducing the time of testing by 80% and improving testing efficiency.

Machine learning/AI intern

🚀 Passionate Machine Learning Engineer | GenAI & NLP Enthusiast | Builder of Scalable AI

Hey there! 👋 I’m a hands-on Machine Learning Engineer with 6+ years’ experience turning cutting-edge AI concepts into real-world, production-ready solutions. Whether it’s designing predictive models that power fintech analytics, crafting GenAI/NLP pipelines for healthcare, or building high-throughput MLOps systems for SaaS, I thrive at the intersection of innovation and impact.

🎯 What I do:

- Architect and deploy robust, scalable ML pipelines that fuel business decisions—using the latest in BERT, LLMs, LightGBM, and deep learning.
- Bring structure to chaos by building automated, CPU-optimised solutions that handle large, fast-moving data for real-time results.
- Power up GenAI and NLP: from prompt engineering to large language model distillation, efficient model training, and feature-rich data engineering.
- Lead as a Kaggle Competition Master 🏅, excelling at solving diverse ML challenges at an expert level.

🛠️ My toolkit:

Languages: Python | C/C++ | NodeJS
Data Science: Machine Learning | Deep Learning | Predictive Analytics | MLOps | Generative AI | NLP
Frameworks: TensorFlow | PyTorch | XGBoost | LightGBM | Scikit-Learn | Transformers | Lang Chain
DevOps & Cloud: Docker | AWS (ECS, EC2, SageMaker, RDS) | Airflow | MLflow | DVC | CI/CD
API & Backend: FastAPI | Flask
Databases: MongoDB | PostgreSQL | Elasticsearch

💡 I love collaborating with driven teams and translating business challenges into smart, scalable, and efficient AI products. From automating complex workflows that cut down manual effort, to boosting performance and accuracy on large-scale platforms, I don't just build models — I unlock value.

Curious about all things GenAI, open to tough datasets, and always ready for the next challenge. Let’s connect and see where we can innovate together! 🤝

taplio

The Battle

Abhishek Thakur

Urvish Patel

Followers

Posts

Engagements

Likes

Ready to grow your LinkedIn presence?