Jad DELLELChief Data & Sustainability Officer| Ivy Decarb
I've been using NotebookLM since it's beta version and I am using its newest plus version, it's an amazing tool, it handles up to 300 documents at once (plus version) it is helping me with industry benchmarking and analysis : my latest use case is uploaing +100 sustainability reports from different industries and extract best practices and highlights by industry.
I use also Gemini which is very helpful in the context of Google Suite ( Sheets, docs,...) , creating sheets, adding formulas, analyzing ...
Perplexity pro for deep search and sometimes compare answers, switching between models is amazing ( Claude, Gemini, Chatgpt 4-o, sonar,...)
Deepseek for different perspectives, concise answers and mathematical related questions
I use ChatGPT for content creation and brainstorming
Saving Changes...
Anonymous
for research related work, I find Claude , elicit and perplexity useful. For Project management ChatGPT and Microsoft co-pilot as well as PMI Infinity useful Saving Changes...
I can use only a specific tool to company restrictions but I have used it for text summarization, document creation, and validating the accuracy of certain datasets. Saving Changes...
Sanjay SinghProject Manager / Business Process Improvement GuruMaharashtra, India
Hello Claudia,
Generative AI is a rapidly evolving field, and having the right tools can make all the difference. Here’s a breakdown of my go-to resources and workflows:
1. Tools for Working with Generative AI Data. Synthetic Data Generation: Tools like Gretel.ai https://gretel.ai/, Mostly AI https://mostly.ai/, and Synthetic Data Vault (SDV) https://sdv.dev/ are fantastic for creating high-quality synthetic datasets when real data is scarce or sensitive. Fine-Tuning & Deployment: I rely heavily on frameworks like Hugging Face Transformers https://huggingface.co/ for model fine-tuning and Weights & Biases (W&B) https://wandb.ai/ for experiment tracking. For deployment, FastAPI https://fastapi.tiangolo.com/ and Streamlit https://streamlit.io/ are my top choices for building scalable APIs and interactive demos.
2. Data Cleaning & Preprocessing Automation: Libraries like Pandas https://pandas.pydata.org and Dask https://dask.org/ handle large-scale data wrangling, while OpenRefine https://openrefine.org is excellent for manual cleaning. LLM-Assisted Cleaning: I’ve started using OpenAI’s GPT-4 or Anthropic’s Claude to help with data labeling, anomaly detection, and generating cleaning rules, saving tons of time!
One underrated practice is "data-centric AI" spending more time refining datasets rather than just tweaking models. Tools like Snorkel AI https://snorkel.ai for programmatic labeling and Cleanlab https://cleanlab.ai for finding label errors have been game-changers in my projects.
I’d love to hear if you’ve experimented with any of these or have alternative tools to recommend, always looking to learn from the community. Saving Changes...
Ashok GuruPLM Manager| Sconce Solutions USA Inc.Atlanta, Ga, United States
I have used GenAI only from an experimental basis for data preparation for data migrations and hence cannot recommend any applications until much more investigations performed. I have tested Chat GPT, Claude and Bard during this exercise and actually enjoyed working with Claude. Saving Changes...
Suresh RamisettiProject Manager| PPSSchaumburg, IL, United States
Dec 08, 2023 2:29 AM
Replying to Hakam Madi
...
Hello Claudia,
I've been experimenting with Perplexity in my AI data, but I can't fully recommend it yet since my experiments are incomplete. However, so far, I like it.
As for charting and visualization tools, if you're building a webpage (internally or externally), I recommend using eCharts and DataWrapper. They're both free and easily accessible at the following links:
If you prefer connecting your data to a data analytics solution, the best advanced-solution is Google Data Studio (Looker Studio). It's free to use and there are plenty of resources available to help you learn how to use it seamlessly.
I find a combination of GenAI tools and learning resources indispensable for enhancing my capabilities. Tools like ChatGPT and Microsoft Copilot help streamline writing, summarization, brainstorming, and documentation tasks. For visual design and content creation, platforms like Canva with AI features and Adobe Firefly are extremely helpful. I also rely on Notion AI for productivity and knowledge organization. Saving Changes...
Microsoft Power BI with Copilot: Enables natural language queries to generate reports and visualizations using GenAI. Tableau, Power BI, Matplotlib, Plotly: For creating dashboards and visualizations to interpret model outputs and performance. Saving Changes...
Copilot and Notebook LM have been good resources. Adobe's AI also has some upside. Saving Changes...
RICHA VYASRICHA VYASMorrisville, NC, United States
We are experimenting with MS Copilot Studio for our work, creating agents for different work and we are in the experiment/testing phase. Saving Changes...
PMI Team Member
