Poonam Solanki

Data Scientist

Pune, Maharashtra, India0 mo experience

AI EnabledAI ML Practitioner

Key Highlights

Developed advanced LLM-powered applications.
Created efficient NLP pipelines for real-time processing.
Integrated AI solutions for document intelligence.

Stackforce AI infers this person is a skilled Data Scientist specializing in AI and Machine Learning applications.

Contact

Skills

Core Skills

Machine LearningNatural Language Processing (nlp)Document IntelligenceComputer VisionWeb Development

Other Skills

.NET FrameworkAIMLAPI DevelopmentAPI ManagementASP.NET Web APIArtificial Intelligence (AI)Azure Cognitive ServicesAzure Document IntelligenceBlazorBootstrapBusiness RequirementsC (Programming Language)C#C# in .NETC++

Experience

0 mo

Total Experience

Average Tenure

Current Experience

Alliancebernstein

Software Development Associate

Jun 2023 – Present · 3 yrs · Pune, Maharashtra, India · On-site

Developed RAG-based chatbot using ChromaDB, FAISS, Gist embeddings, and prompt engineering (Zero-shot, CoT).
Evaluated LLMs and RAG pipelines using TruLens, DeepEval, PredictionGuard, and RAGAS for performance benchmarking.
Built ML model (Logistic Regression) to classify AI vs human content; used PCA, ROC-AUC, precision, and recall for analysis.
Created NLP pipelines for news summarization and intent recognition using LangChain, spaCy, and NLTK.
Integrated Azure Document Intelligence and PyTesseract for OCR and document data extraction.
Worked with LLaMA and Gemma models for fine-tuned generative AI use cases.
Built Streamlit tools for tokenization and PDF summarization with LLM evaluation workflows.
Automated email alert systems and chatbot (LUIS) to streamline internal processes and reduce manual effort.
Developed a CGM Rule Testing Framework to automate compliance checks and reduce trading errors.
Tech Stack: Python, SQL, PyTorch, TensorFlow, Azure Cognitive Services, PyTesseract, LLaMA, Gemma, ChromaDB, FAISS, LangChain, Streamlit, GitHub, RESTful APIs, spaCy, NLTK, Scikit-learn

PythonSQLPyTorchTensorFlowAzure Cognitive ServicesPyTesseract+13

Tata motors

Project Intern

Jan 2023 – May 2023 · 4 mos · Pune, Maharashtra, India · On-site

🔹 AI/ML & Computer Vision
Developed an AI-powered object detection and tracking system using YOLO algorithms, improving accuracy in live video streams.
Processed 15,000+ objects daily, reducing manual counting errors by 60-70% and enhancing operational efficiency.
🔹 Automation & Performance Optimization
Automated real-time object tracking and classification, significantly reducing manual effort and human errors.
Enhanced workflow efficiency by 30%, leading to improved vendor accountability and cost savings.
🔹 Cost Reduction & Business Impact
Saved 75+ labor hours per week by automating object detection and reducing dependency on manual monitoring.
Reduced manpower costs by 5-10%, improving overall cost efficiency.
🔹 Tech Stack & Tools Used
AI/ML & Computer Vision: YOLO, Python, OpenCV
Data Processing: Pandas, NumPy
Automation: Azure Cognitive Services, Image Annotation Tools

YOLOPythonOpenCVPandasNumPyAzure Cognitive Services+2

The akshaya patra foundation

Intern

Sep 2021 – Nov 2021 · 2 mos · Pune, Maharashtra, India

🔹 Website Development & Fundraising
Designed and developed a philanthropic website to raise funds for underprivileged children, improving online donor engagement.
Helped raise ₹18,000, increasing student beneficiaries by 20% and boosting the NGO’s fundraising by 15%.
🔹 Payment Integration & Transparency
Implemented secure UPI payments to streamline donation processing and enhance contributor trust.
Developed a transparent tracking system to provide real-time updates on fund utilization.
🔹 Tech Stack & Tools Used
Web Development: HTML, CSS, JavaScript
Payments & Security: UPI Integration, Secure Payment Gateways

HTMLCSSJavaScriptUPI IntegrationWeb Development

The sparks foundation

Web Developer

Aug 2021 – Sep 2021 · 1 mo

Jpmorgan chase & co.

Intern

Aug 2021 – Aug 2021 · 0 mo

In this virtual internship i set up a system so that it can interface with the relevant financial data feed, made the required calculations and then presented this in a way that allows the traders to visualize and analyze this data in real time.