Data First Consultancy @DataFirstGroup

We empower businesses to harness the full potential of their data through expert-led strategy, services and products datafirstconsultancy.co.uk London, United Kingdom Joined August 2025

Tweets

111
Followers

238
Following

20
Likes

32

Data First Consultancy @DataFirstGroup

14 hours ago

Just rolled out a real-time Metabase dashboard for a client - open-source, no-code, up in hours. Fancy painless BI? Have a nose at the write-up. #Metabase #OpenSourceBI #DataAnalytics blog.datafirstconsultancy.co.uk/p/unlocking-bu…

0 0 0 20 0

sid @immasiddx

a day ago

Don’t worry, our jobs are safe.

375 693 19K 703K 845

Download Image

Data First Consultancy @DataFirstGroup

a day ago

OpenAI’s latest note digs into hallucinations - they pin much of the blame on weak evals and misaligned objectives. Helpful to see candid metrics, though it still skirts the messy incentive issues when these models hit production. Worth a read. #AI #LLM openai.com/index/why-lang…

0 0 0 26 0

Data First Consultancy @DataFirstGroup

a day ago

Airbyte's write-up on its certified ClickHouse destination gets into batching and ReplacingMergeTree to outpace the old connector, but it's light on latency or cost data - run your own benchmarks before switching. #ClickHouse #ELT #DataOps airbyte.com/blog/clickhous…

0 0 0 25 0

Data First Consultancy @DataFirstGroup

a day ago

Snowflake’s new guide bundles Posit Workbench + Connect as a native app. Handy single stack: open source dev, one-click Shiny/Streamlit deploy, inherited roles. Worth noting the trade off: less infra pain, more Snowflake lock in. #DataScience #LLM medium.com/snowflake/from…

0 0 0 23 0

Data First Consultancy @DataFirstGroup

a day ago

Skimmed Airbyte’s take on Airbyte vs Airflow. Good distinction: Airbyte ships ELT connectors, Airflow orchestrates jobs. Fair, but they skip over teams that pair both for lineage, retries and heavy DAG logic. #dataengineering #ETL airbyte.com/blog/airbyte-v…

0 0 0 23 0

Data First Consultancy @DataFirstGroup

a day ago

Snowflake's latest builders blog pitches an LLM-powered Slack bot to surface buried team knowledge. Smart architecture (embeddings, confidence gates), yet success still hinges on disciplined Slack hygiene, accurate tagging and budget for Cortex calls. #AI #KnowledgeManagement…

0 0 0 21 0

Data First Consultancy @DataFirstGroup

a day ago

Snowflake outlines Adaptive Cortex Complete, an in-warehouse bandit router that chooses the 'best' LLM per prompt. Cost-vs-quality trade-off is neat, but relying on other models to judge answers risks circular bias. Try, measure, verify. #LLM #DataEng medium.com/snowflake/adap…

0 0 0 21 0

Data First Consultancy @DataFirstGroup

a day ago

Fresh read from OpenAI on why language models hallucinate; they pin the issue on weak evals and misaligned incentives, not model size. Tighter benchmarks are welcome, but the paper slides past messy data quality and hidden system prompts - the very roots of hallucination.…

0 0 0 48 0

Data First Consultancy @DataFirstGroup

a day ago

DE Weekly #234 shows data work stretching in three directions at once. 1) Real time: Dagster engineers outline wiring Kafka and Flink into event driven pipelines with second level observability. 2) Unstructured AI: Netflix and Etsy move from metrics to video, images and…

0 0 0 19 0

Data First Consultancy @DataFirstGroup

a day ago

Zero-heavy metrics like purchase counts can fool a vanilla Poisson. Arnaud Capitaine shows how to test for zero inflation and why a hurdle or ZINB model often tells the story better. Handy guide before you blame 'overdispersion'. #DataScience #Statistics towardsdatascience.com/zero-inflated-…

0 0 0 16 0

Data First Consultancy @DataFirstGroup

a day ago

Just read Airbyte’s side-by-side of SQL and Python for analysis. Sensible rundown: SQL shines for set-based ops & governance, Python for advanced stats & automation. They rightly note team skills trump tool debates, yet gloss over how hybrid workflows complicate lineage &…

0 0 0 14 0

Data First Consultancy @DataFirstGroup

a day ago

Not your typical ‘how we 10x-ed growth’ post - RudderStack explains how Magic Eden seized 95% of Solana NFT volume by funnelling every customer event into a warehouse-first setup, then pushing cleansed data back to product and marketing. Solid proof of event streaming + reverse…

0 0 0 15 0

Data First Consultancy @DataFirstGroup

2 days ago

Snowflake’s latest piece pits LAF against Databricks Delta Sharing + Deep Clones, claiming 50% lower cost and much less plumbing. Sensible assumptions, but based on Large WH pricing and a not-yet-GA Iceberg path. Worth running your own TCO. #DataOps #CloudData…

0 0 0 17 0

Data First Consultancy @DataFirstGroup

2 days ago

Quick refresher for anyone prepping data interviews: joins, window functions, aggregation logic and NULL quirks still trip up plenty of us. Nate Rosidi outlines six pitfalls with clear fixes - worth a read before your next panel. #SQL #DataCareers kdnuggets.com/most-candidate…

0 0 0 15 0

Data First Consultancy @DataFirstGroup

2 days ago

Databricks outlines AI-powered journeys: vector search and personalised recs on a lakehouse to lift sales. Sensible, but it assumes pristine 1st-party data and solid MLOps - not a quick win for every retailer. Build vs buy is still open. #ecommerce #AI databricks.com/blog/search-sa…

0 0 0 16 0

Data First Consultancy @DataFirstGroup

2 days ago

Snowflake’s fresh blog unwraps the Gen AI specialty cert, opening with Cortex Search. Good tips on chunking, TARGET_LAG and credit watching, but bear in mind it all stays in the Snowflake garden. #Snowflake #GenAI medium.com/snowflake/%EF%…

0 0 0 16 0

Data First Consultancy @DataFirstGroup

2 days ago

Muhammad Ardi walks through MobileNetV1 from scratch in PyTorch. Good refresher on how depthwise separable convs cut >80% compute, and why width / res multipliers often beat stacking layers. Handy for edge deployments. #DeepLearning #PyTorch towardsdatascience.com/the-tiny-giant…

0 0 0 13 0

Data First Consultancy @DataFirstGroup

2 days ago

Useful rundown from Erika G. Gonçalves on why GenAI apps need the same discipline as traditional software. Continuous eval loops, clear acceptance criteria, and auditable pipelines aren’t optional if you want anything beyond a flashy demo. #MLOps #GenAI towardsdatascience.com/ai-operations-…

0 0 0 15 0

Data First Consultancy @DataFirstGroup

2 days ago

OpenAI spotlights Blue J, a GPT-4.1 tool for tax research. Cited answers are handy, but AI reasoning still hides behind a black box - regulators and CFOs will ask how conclusions were drawn. A useful aid, yet not a rubber stamp. #AI #taxtech openai.com/index/blue-j