Not your typical happy hour. ?? During the week of Data Council, join us for Data Renegade—where the rebels of data engineering gather to trade bold ideas, share hard lessons, and connect over good drinks and better company. Hosted by Recce - Making Data Productive., Tobiko and Datacoves, this is your space to meet fellow builders, challenge the status quo, and spark real conversations—one table at a time. ?? RSVP now: https://lu.ma/ctsozuun #DataCouncil #DataEngineering #DataCommunity #HappyHour #DataRenegade
Recce - Making Data Productive.
数据基础架构与分析
Helping data teams preview, validate, and ship data changes with confidence.
关于我们
Recce helps modern data teams preview, validate, and ship data changes with confidence. By turning pull requests into structured, context-rich reviews, Recce makes it easy to spot meaningful changes, verify intent and impact, and reduce cognitive load for authors and reviewers alike. Curate reproducible checklists that compare data across environments — so you can catch what matters, skip what doesn’t, and align your team before merging. Accelerate development, cut down on manual QA, and bring visibility, verifiability, and velocity to your data workflows.
- 网站
-
https://datarecce.io
Recce - Making Data Productive.的外部链接
- 所属行业
- 数据基础架构与分析
- 规模
- 2-10 人
- 总部
- San Francisco
- 类型
- 私人持股
- 领域
- dbt、Modern Data Stack、code review、Data Engineering、SQL、Data Lineage、Query Diff、Lineage Diff和Data Model Diff
地点
-
主要
获取路线
US，San Francisco
Recce - Making Data Productive.员工
-
Udit Panjwani
Ingenious Filmmaker | Disruptive Industry Expert | Social & Digital New Media | Visual Effects Supervisor
-
Dave Flynn
DataRecce.io | Developer Relations, Marketing, Developer Outreach
-
Dori Wilson
YC Alum | Interim Head of Data @ Recce | Data, Startups, & Coffee Chats
-
Karen Hsieh
Data ???Empower ???Product
动态
-
Kick Off Data Council with Data Reboot: Lightning Talks + Happy Hour Let’s reboot before the big day. Join us on April 21st for Data Reboot—an evening of lightning talks, lively discussions, and casual hangs with the data community. Whether you're into data engineering, AI, or analytics, you'll find fresh ideas, bold perspectives, and great company. Talks are short, fun, and inspiring—just enough to spark the vibe before the conference. ?? Near the Data Council venue. ?? April 21st, 6:00–9:00 PM https://lu.ma/ro29lcyb See you there to kick off Data Council 2025 right! ??
-
Recce is officially SOC 2 Type 1 compliant! At Recce, keeping your data secure and reliable is fundamental. Achieving SOC 2 compliance is a meaningful step, validating that our security, availability, and confidentiality controls meet rigorous industry standards. For our users this means that Recce delivers actionable visibility, verifiability, and velocity for your data changes with your data security at the core of that. This milestone reaffirms our commitment to helping data teams confidently contextualize and validate data impacts during development and pull-request reviews. Find details in our Trust Center:?https://lnkd.in/ggq_5fiD Read more on the Recce blog: https://lnkd.in/gs5_24uM Let’s continue working together to empower data teams to merge faster, safer, and with greater clarity. #SOC2 #DataSecurity #DataEngineering #OpenSource #TrustAndCompliance
-
-
Using Column-Level Lineage to Validate dbt Data Model Changes - save time and increase accuracy of your data validation work We recently added column-level lineage to Recce after recognizing the value it brings to data validation workflows in dbt projects. In the blog post below, we walk you through how column-lineage can help you speed up and improve the accuracy of your QA work for PRs in your dbt project: https://lnkd.in/gymBbP8f #dbt #dataengineering #dataquality #datavalidation #columnlineage #opensource #Recce #dataimpact #analytics
-
-
see how our user adopts the reusable checklist to share your works and knowledge. ??
?? Optimise ta validation de données avant MEP avec une checklist réutilisable sur Data Recce. --- ??? Construire une checklist 1?? Lancer un Recce Server et charger les artefacts dbt (target et target base). 2?? Explorer les changements dans le lineage avec Explore Change et ajouter des checks standards (row count, value diff…). 3?? Créer des checks personnalisés dans l’onglet Queries en écrivant tes propres requêtes. 4?? Ajouter chaque test à la checklist en cliquant sur Add Checklist. 5?? Organiser et documenter la checklist en renommant et ordonnant les checks. --- ?? Exporter la checklist Une fois construite, exporte-la en JSON grace au bouton Export ou Save. Ce fichier contient : ? La liste des checks et leurs résultats. ? Les artefacts dbt (targets). ? Des métadonnées utiles à l’analyse. --- ?? Réutiliser la checklist Une fois exportée, elle peut être rejouée de plusieurs fa?ons : Afficher une synthèse des erreurs ? recce summary my_recce_state.json Analyser les résultats sans nouvelle exécution ? recce server --review my_recce_state.json Relancer les tests depuis l’interface web ? recce server my_recce_state.json Exécuter tous les checks en ligne de commande ? recce run --state-file my_recce_state.json --- Bonus : Partage ce fichier avec ton équipe pour collaborer et enrichir l’analyse. Avec cette approche, la validation devient plus rapide, reproductible et évite les allers-retours inutiles. ?? Tu utilises déjà Data Recce ? Partage ton expérience !
-
-
Data Reboot: Lightning Talks & Happy Hour Ahead of Data Council Get a head start on the conversations, connections, and ideas of Data Council with good food, great company, and rapid-fire lightning talks at Data Reboot on April 21st! This is your chance to meet fellow data practitioners, spark discussions, and warm up for the conference in a relaxed, friendly setting. Expect fun, fast-paced talks on data engineering, analytics, AI, and beyond—fresh takes, bold ideas, and insights too good to miss. ?? Where: Office space near the conference venue ?? When: April 21st, 6:00 - 9:00 PM https://lu.ma/ro29lcyb
-
-
Column-level lineage in Recce 0.57 helps you work smarter by tracking data evolution across your dbt project - target models for data validation with speed and precision Column-lineage is invaluable for drilling down and checking for impact, and aligns with Recce's methodology of streamlining your checks following data model updates. 1?? Lineage Diff?- shows only potentially impacted models 2?? Breaking Change Analysis?- eliminates model changes that would not impact data 3?? Column-Level Lineage?- shows the evolution of a column as it propagates downstream Each step progressively helps you further drill-down into the what-and-why of your data modeling changes. Try it now in version 0.57: pip install -U recce It’s amazing how this can speed up your validation work, while still retaining complete and comprehensive coverage of data impact. Let us know when you try it out! #data #datamodeling #datavalidation #dataengineering #analytics #analyticsengineering #sql #lineage #columnlineage #dbt
-
-
?? Missed the Data Council talk submission deadline? Here’s your second chance! ?? Recce is hosting a community Welcome Dinner + Lightning Talks on April 21st (Day 0) of Data Council 2025, and we’re looking for speakers to share 5-minute lighting talks! It’s an opportunity to connect with people before the main conference kicks off. If you have a bold idea, fresh perspective, or exciting insight that data practitioners would love, we want to hear from you. Topics can be data engineering, analytics, AI, scaling data teams, or anything that sparks curiosity in the data community. ?? Where: Office space near the conference venue ?? When: April 21st, 6:00 - 9:00 PM ? Talk format: 5 minutes per speaker (slides optional) ?? Spots are limited, so apply early! ?? Interested? Fill out this quick form: https://lnkd.in/gabvqVYJ to submit your topic by Apr 10th. Let’s warm up for the Data Council 2025 with fresh ideas and great conversations! ???DM me if you have any question. #DataCouncil2025 #DataCommunity #LightningTalks #CallForSpeakers
-
still writing ad-hoc queries? see how others do?
?? Personnaliser la validation des données avec DataRecce La validation des données avant MEP repose sur deux approches : ? Analyse générique : Comptage des lignes, valeurs distinctes par colonne, profiling classique. ? Analyse spécifique : Requêtes ad hoc adaptées aux évolutions métier et aux données. Pour cette seconde approche, on construit souvent une requête SQL de comparaison entre environnements. Un exercice techniquement intéressant, mais vite répétitif et chronophage. --- ?? Exemple concret Comparer les montants par client entre prod et préprod après une évolution sur les produits d'une catégorie A. La query de comparaison est : >> WITH target_query_preprod AS ( SELECT id, SUM(mt) AS total FROM my_model WHERE cat = 'A' GROUP BY id ), base_query_prod AS ( -- Query similaire ), all_records AS ( SELECT * FROM base_query_prod UNION ALL SELECT * FROM target_query_preprod ), differences AS ( -- Comparaison des écarts entre environnements -- query expression à écrire ) SELECT * FROM differences; --- Avec DataRecce, les choses sont plus simples. Dans l’onglet Query, il suffit de saisir la requête target_query_preprod et de définir la clé unique 'id'. Le bouton Run Diff génère et exécute automatiquement la requête complète et affiche directement les différences : En vert les clés qui seront ajoutées après MEP En rouge celles qui seront supprimées En blanc les clés identiques avec des valeurs différentes Résultat : un gain de temps énorme et un focus immédiat sur l’analyse des écarts. --- ?? Exemple plus complexe où on souhaite filtrer les écarts pertinents. En prod, on a les ventes des catégories A et B. En préprod, on a ajouté les ventes de la catégorie C. Objectif : Vérifier l'absence d'impact sur A et B. On doit comparer les résultats des queries expressions suivantes : -- base_query_prod >> SELECT id, SUM(mt) AS total FROM my_model GROUP BY id; -- target_query_preprod >> SELECT id, SUM(mt) AS total FROM my_ -- ici exclure la catégorie C WHERE cat IN ('A', 'B') GROUP BY id; --- ?? La feature Custom Queries En activant cette feature, on peut exécuter des queries différentes par environnement. Cette feature partage l’écran en deux panneaux : Base pour la prod et cible pour la préprod (Voir l'image d'illustration). On peut personnaliser les filtres directement dans l’interface. Et en cliquant sur Run Diff, on peut visualiser les écarts. --- En résumé, pour faire une analyse spécifique, l'onglet Query centralise tout et simplifie la validation des données. Il offre en bonus d'autres features : L’historique des runs pour rejouer ou analyser des tests passés sans recréer les requêtes. Une checklist de validation exportable en JSON pour structurer et partager les contr?les. Une capture automatique des résultats sous forme de tableau pour faciliter le reporting dans une merge request. Si tu utilises déjà DataRecce ? Partage ton retour d’expérience !
-
-
Are you using column-level lineage yet? There’s no denying the value for performing reconnaissance on your data project, but the differences between each implementation are more than you’d think. Read the article: https://lnkd.in/gm69i6tW We took a look at the main column-level lineage options that are built for the dbt data transformation platform: - Power User for dbt - dbt Explorer (dbt Cloud) - SQLMesh (with dbt compatibility) Surprisingly, there are some key differences that help to infer the intended use case for each implementation. dbt column level lineage in Explorer is designed for understanding a column before you make changes to your project. In SQLMesh, only upstream column evolution is shown, implying that the feature is for root cause analysis. How about transformation code? - Would you like your CTEs to be nodes in the column-level lineage? - Would you like each column to in the evolution to be a separate node? Find out: https://lnkd.in/gm69i6tW #dbt #SQLMesh #Data #SQL #Analytics #Lineage #DataLineage #ColumnLineage #ColumnLevelLineage #DataEngineering
-