~/talks

Pietro Mele

what they talk about

all 8 in the filters

where they keep going back

all 8 in the filters

on a stage with

all 5 in the filters

who talks about the same things

titleconferencedate
The Real Cost of a Token: The Hidden Bill of Generative AIVoxxed Days Luxembourg 20262026-06-18
Examines the real monetary and compute costs of tokens in generative AI, naming models and inference/serving infrastructure as the main cost drivers. Explains…
Talk in French on
Reviving phonetic algorithms for better search relevanceBerlin Buzzwords 20262026-06-07
Reimplements century‑old phonetic encoders and benchmarks them against Levenshtein fuzziness on a huge French audiovisual index, finding phonetic matching cuts…
Talk on
The Real Cost of a Token: the Hidden Bill of Generative AIDevoxx France 20262026-04-22
The talk quantifies the environmental cost of generative AI beyond token or latency metrics, measuring energy (kWh), water use and e‑waste from model training…
Talk in French on
Unlocking Next-gen Semantic Search With OpenSearch 3.xOpenSearchCon Europe 20262026-04-16
OpenSearch 3.x demonstrates concrete recipes for billion‑scale vectors: GPU indexing via Cagra, Faiss and JVector engines, and HNSW/IVF ANN with Radial Search…
Talk on
Behind the hype: managing billion-scale embeddings in Elasticsearch and OpenSearchVoxxed Days CERN 20262026-02-10
Managing 1 billion embeddings in Elasticsearch and OpenSearch requires concrete tradeoffs: GPU + NVIDIA Cagra speeds up index building but a naive float32 ANN…
Talk on
A Beginner’s Guide to Semantic SearchVoxxed Days CERN 20262026-02-10
Compares classic TF-IDF and BM25 lexical search with dense neural embeddings and shows concrete tradeoffs: lexical can be tuned but fails on vocabulary…
Talk on
Building Next-gen Search With OpenSearchOpenSearchCon North America 20252025-09-08
OpenSearch 3.x exposes agentic RAG: register a model (connector + model group), create a plan-execute-reflect agent with conversation index memory, and bind…
Talk on
Hybrid search on hybrid models, at scaleBerlin Buzzwords 20252025-06-15
Extensible hybrid search runs on Elasticsearch multi-index architecture with OpenVINO inference, Spring Boot microservices and Kubernetes; Whisper…
Talk on
Exploring Large Graphs at the Heart of the French National Audiovisual Ins.Big Data & RPA Conference Europe 20242024-11-19
French INA metadata model uses a cultural‑heritage graph and powers ~60 million documents stored as relational rows, MongoDB fragments and Elastic indexes…
Talk on
Home Assistant under surveillanceVoxxed Days Luxembourg 20242024-06-20
Home Assistant monitoring: how to keep a Home Assistant smart-home installation observable and reliable. Covers technologies for collecting and visualizing…
Talk in French on
Implementing an open-source RAG with OpenSearchOpenSearchCon Europe 20242024-05-06
An open-source RAG pipeline is implemented using Langchain for orchestration and OpenSearch as the vector store, with embeddings stored as 384-dimension…
Talk on
Building Your Intelligent Assistant with Hugging Face and ElasticsearchDevoxx France 20242024-04-17
Shows a concrete pipeline that crawls and indexes about 2,900 Devoxx talks into Elasticsearch, computes 1536‑dim dense embeddings plus sparse embeddings, and…
Talk in French on
Set Up an Open Source RAG in 30 MinutesDevoxx France 20242024-04-17
Shows building a complete open-source RAG pipeline in minutes using LangChain for prompt construction and inference orchestration, OpenSearch 2.8 connectors to…
Talk in French on
Cloning ChatGPT with Hugging Face and ElasticsearchDevFest Toulouse 20232023-11-16
Shows how to combine Hugging Face models and Elasticsearch to build powerful data-augmented retrieval and inference applications. Demonstrates prototypes that…
Talk in French on
Cloning ChatGPT with Hugging Face and ElasticsearchVoxxed Days Luxembourg 20232023-06-21
Elasticsearch (a search engine, NoSQL datastore and analytics engine) and Hugging Face (an open-source ML model platform) are combined to build powerful…
Talk in French on
~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.