Skip to content
View roberto-ceraolo's full-sized avatar
  • Lausanne

Highlights

  • Pro

Block or report roberto-ceraolo

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
roberto-ceraolo/README.md

Hello! I'm Roberto Ceraolo, thanks for being here!

I'm Italian, 24 years old, curious and energetic!

I am captivated by human ingenuity and the idea of empowering it with technology.

I am a Data Science student at EPFL in the beautiful Lausanne, Switzerland ⛰️

Present

Currently, I am at ETH in Zurich, working with Zhijing Jin and Mrinmaya Sachan on the intersection of Causal Inference and AI agents.

Here you can find some of my thoughts on the future of AI agents.

On occasion, I help startups, either on technical aspects, or on strategic decisions.

Future

I am looking for a full time position starting December 2024 or January 2025.

If you’re looking for someone that

  • has technical expertise in AI but also cares about the business side,
  • is able to prioritize and put things in a customer-centric perspective,
  • loves the 0 to 1 of building a product, then I might be your guy.

Some projects I've worked on recently:

I built Sentence Pair Finder, a system to find pairs of sentences which are similar in meaning but diverse in words in large corpora.

I worked on CausalQuest, a dataset of Natural Causal Quesitons (Paper under review)

🔭 I worked on building a vector search retrieval system and generative models for Gamma Ray Logs (Earth subsurface measurements) at Schlumberger-Doll Research, Cambridge, MA. (Patent pending approval)

Fine-tuning an AI Assistant for STEM Education

Development of an AI assistant specifically suited for STEM students. This included data gathering and preparation, training of a reward model (Deberta), fine-tuning of a large language model (Distilled GPT-2). Used the Antrophic-style Constitutional AI approach in order to bias the model towards providing clear, correct, complete, and rigourous answers.

Diffusion Models for Video Panoptic Segmentation

Re-implementation of the whole codebase of Pix2Seq-D, a diffusion-based model to approach segmentation tasks, in Pytorch from Tensorflow, and adapted of the architecture to the task of Video Panoptic Segmentation.

OpenRitardi

Development of a web app providing visualizations and search functionalities on train delays in Italy. The goal is to spread awareness about the quality of the service and prompt the company handling the railway system, to publish their data.

Causal Representation learning

Study with the goal of improving generalization capabilities of deep learning algorithms. Investigated the usage of Independence Regularization in Causal Representation Learning to increase the accuracy of a classifier on out-of-distribution data.

Before saying goodbye

You can find more projects I've worked on and my full cv in my personal website.

If you want to connect with me, feel free to send me a DM on Linkedin!

Fun fact: I am a keen believer of Parkinson's Law

Pinned Loading

  1. giacomoorsi/OpenRitardi giacomoorsi/OpenRitardi Public

    OpenRitardi is the first visualization tool of train delays in Italy. You can see the performance of each station and train, and see where the railway service can be improved. The project showcases…

    Jupyter Notebook 16 2

  2. causal-quest causal-quest Public

    CausalQuest: Collecting Natural Causal Questions for AI Agents

    Python 6

  3. semantic-diverse-search semantic-diverse-search Public

    A system to find pairs of sentences which are similar in meaning but diverse in words in large corpora.

    Python

  4. frasalvi/lauzhack2022 frasalvi/lauzhack2022 Public

    Collaborative platform where EPFL students can suggest improvements for the sustainability of the campus

    HTML 2