Accenture

Senior Data Engineer

Ocelot Consulting

• Designed and orchestrate a robust data ingestion pipeline to efficiently pull over 50 million records of data from Teradata and Oracle databases into an S3 bucket.
• Built a scalable MLOps pipeline using Amazon SageMaker, which involves data preparation, model training, deployment, and monitoring.
• Implemented monitoring and logging mechanisms to track model performance, system health, and data quality, enabling proactive troubleshooting and continuous improvement.

MLOps Engineer

Prefect

• Assisted over 50 data engineers and data scientists in building efficient data pipelines that incorporated best engineering practices and tools. 
• Wrote and tested Python code that met 90%+ unit test pass rates, resulting in a more reliable product and increased customer satisfaction.
• Collaborated with cross-functional teams to define software requirements that fulfill customer needs, resulting in a 10% decrease in development time.

Developer Experience Engineer

Developer Advocate

NVIDIA

I write articles related to data science on NVIDIA Developer Blog once per month to spread awareness on data science tools and best data science practices.

Technical Writer

Freight Science

• Built an end-to-end data science pipeline that streamlines data validation, data processing, machine learning training, monitoring, and output visualization.
• Wrote concise and modular code for data workflows, complete with unit tests, resulting in improved accuracy and reproducibility of experimental outcomes.
• Collaborated with stakeholders to deconstruct the business objectives into actionable hypotheses, improving decision-making and driving positive outcomes.

Data Scientist

SIUE Tutoring Resource Center

• Guided more than ten students to get from C to A grades in statistics and mathematics courses within two months. 
• Provide small group tutoring sessions in subjects including Algebra, Calculus, Geometry, Statistics, and Probability.

Math and Statistics Tutor

Workera

• Design and implement 100+ Python-based assessments for data scientists and data engineers.
• Focus on evaluating the knowledge of Python functions, classes, variables, and best practices for OOP in the assessments.

Python Assessment Developer

Datapane

• Boosted monthly views by 1200+ by creating engaging example analyses and templates in Datapane's documentation and gallery.
• Developed compelling data science content and examples.
• Collaborated with domain experts, data scientists, and backend engineers to create innovative features for Datapane.

Data Science Evangelist

Saturn Cloud

Contributor

Universidad de Guanajuato

Collaborative research with Dr. Adrián Pastor López in predicting the author’s gender and language variety by processing raw Twitter data, applying NLTK, Scikit-learn, and experimenting with different NLP models. https://github.com/khuyentran1401/Twitter

Undergraduate Research Assistant

I love exploring and sharing data tools that empower data scientists to work more productively. Subscribe to my newsletter at https://codecut.ai/

Remove image backgrounds locally in 2 lines of Python code 🪄

Removing backgrounds from images typically requires Photoshop, online tools, or AI assistants like ChatGPT.

But these options come with subscription costs, upload limits, or privacy concerns with your images on external servers.

Rembg uses AI models to remove backgrounds locally with just 2 lines of Python.

Plus, Rembg is open source! Install it with "pip install rembg". 

☕️ Run this code: https://bit.ly/48lIIWc
🔗 Link to the repo: https://bit.ly/4ofG9Lj 

#Python #MachineLearning #ImageProcessing #DataScience

Khuyen Tran

Author of Production-Ready Data Science | DevRel @ Nixtla

I recently found out that my role at Nixtla was impacted by company restructuring. It was a shock, but I am excited for what comes next, and I am now open to new opportunities.

I am looking for a remote Developer Advocate or Engineer role in data science, data engineering, or AI. I am fully authorized to work and do not require visa sponsorship.

What I bring:
• 3 years as a developer advocate in open source startups in the data space (Prefect, Nixtla)
• 3 years as an MLOps and data engineer, including experience at Accenture
• Author of Production-Ready Data Science
• Written 200+ technical articles and 1,000+ social posts on Python and data science
• Speaker at PyData and guest on podcasts such as Real Python and Talk Python To Me
• Creator of CodeCut, a site for Python and data science tips with 20K+ monthly views

To see examples of my writing and educational style, you can browse my work at https://codecut.ai

If you know of teams hiring for DevRel or engineering roles in this space, I would love introductions or referrals.

You can reach me on LinkedIn or by email at khuyentran@codecut.ai 

#DataScience #DeveloperAdvocate #DataEngineering #Python

Author of Production-Ready Data Science | Founder @ CodeCut

Auto-annotate PDF images with local vision models (data stays private) 🖼️

Images in PDFs like charts, diagrams, and figures are invisible to search and analysis. Manually writing descriptions for hundreds of figures is impractical.

You could use cloud APIs like Gemini or ChatGPT, but that means API costs at scale and your documents leaving your infrastructure. 

Docling runs local vision language models (Granite Vision, SmolVLM) to automatically generate descriptive annotations for every picture in your documents, keeping data private.

Key benefits:
• Privacy: Data stays local, works offline
• Cost: No per-image API fees
• Flexibility: Customizable prompts, any HuggingFace model 

🚀 Full article: https://bit.ly/4ilsIZ7 
☕️ Run this code: https://bit.ly/3XgRfVt 

#Python #DataScience #MachineLearning #VisionModels

Transform mathematical concepts into engaging animations with Manim ✨

Static slides can only go so far when you're explaining complex concepts.

Dynamic visuals make abstract ideas clearer, more engaging, and easier to understand.

Manim gives you the power to create professional mathematical animations in Python, just like the ones you see in 3Blue1Brown's videos.

In the code below, Manim transforms equations into smooth visual steps:
• Define equation steps using MathTex with LaTeX notation
• Animate equation transformations with the Transform class
• Control animation flow with play() and wait() methods
• Render output with simple command: manim -p -ql script.py

🚀 Full article on Manim: https://bit.ly/49ojNUi 

#Python #DataScience #Manim #DataVisualization

Reduce agent token usage with LLMToolSelectorMiddleware (automated tool filtering) ⚡

Agents with many tools waste tokens by sending all tool descriptions with every request.

This wastes tokens on irrelevant tool descriptions, making responses slower and more expensive.

LangChain v1.0 introduces LLMToolSelectorMiddleware that pre-filters relevant tools using a smaller model.

Key features:
• Pre-filter tools using cheaper models like GPT-4o-mini
• Limit tools sent to main agent (e.g., 3 most relevant)
• Preserve critical tools with always_include parameter 

🚀 Full article: https://bit.ly/4oG7HKx 
☕️ Run this code: https://bit.ly/4r6vWmW 

#LangChain #AI #LLM #Python

Reduce API payload sizes with Pydantic's exclude_computed_fields parameter 🎯

By default, Pydantic's model_dump() serializes computed fields alongside the base fields used to derive them.

This duplicates data and increases API response sizes.

Pydantic v2.12 adds the exclude_computed_fields parameter to model_dump().

This lets you keep computed fields for internal use while excluding them from API responses. 

☕️ Run this code: https://bit.ly/48sHNmU 
🚀 Docs: https://bit.ly/48wV7qz 

#Python #Pydantic #API #DataValidation

Adding 8 hours to 10pm shouldn't give you the wrong morning time, but with Python's datetime, it can.

The standard library fails during DST transitions, returning incorrect offsets when clocks change for daylight saving.

Type hints can't distinguish naive from aware datetimes, leading to runtime bugs.

Whenever eliminates these issues with DST-safe arithmetic and explicit datetime types.

Key benefits:
• DST-safe arithmetic with automatic offset adjustment
• Type safety prevents naive/aware datetime bugs
• Clean timezone conversions with .to_tz()
• Nanosecond precision for deltas and timestamps
• Pydantic integration for serialization

Install it with "pip install whenever".

☕️ Run this code: https://bit.ly/3WSSnyf
📬 Want more Python tips? Subscribe to my newsletter at https://bit.ly/4p4WfZX

#Python  #DateTime #DataScience

Imagine you need to run data processing on a file that is larger than your laptop's RAM. What should you do?

Traditional solutions require buying more RAM, renting expensive cloud VMs, or learning Kubernetes. All of these add complexity and cost.

Coiled's serverless functions let you run your Python code on cloud VMs with the memory you need by simply adding a decorator.

Key capabilities:
• Use any data framework: pandas, Polars, DuckDB, Dask, and more
• Process multiple files in parallel with .map()
• Sync local packages to cloud without Docker
• Cut costs with spot instances and auto-fallback 

🚀 Full article: https://bit.ly/4pFtqCH 
☕️ Link to Coiled: https://fandf.co/4onPt0p 

#Python #DataScience #CloudComputing #sponsored

I had an amazing time speaking at PyData Seattle with Yibei Hu “Multi-Series Forecasting at Scale with StatsForecast!”

We covered how to:
• Use Nixtla’s StatsForecast models with automated model selection
• Apply time series cross-validation for model comparison
• Automate model selection with performance metrics
• Explore TimeGPT for zero-shot and fine-tuned forecasting

Thanks to everyone who joined us!

Watch the talk: https://bit.ly/3LGVCXc 

#PyData #TimeSeries #Forecasting

Handle datasets larger than RAM with DuckDB's automatic spillover 🚀

When datasets exceed available RAM, most tools crash mid-operation.

This forces manual data chunking or expensive hardware upgrades just to complete basic queries.

DuckDB automatically spills intermediate results to temporary files when data exceeds configured memory limits.

Key benefits:
• Process datasets larger than RAM without code changes
• Configure memory limits to prevent system crashes
• Automatic disk spillover when memory fills
• No manual chunking or batching required 

🚀 Full article: https://bit.ly/4ovyUir 
☕️ Run this code: https://bit.ly/4ovyUir 

#Python #DuckDB #DataScience #DataEngineering

Optimize Polars queries with programmatic expression generation ⚡

When you want to use for loops to apply similar transformations, each Polars with_columns() call processes sequentially.

This prevents the optimizer from seeing the full computation plan.

Instead, generate all Polars expressions programmatically before applying them together.

This enables Polars to:
• See the complete computation plan upfront
• Optimize across all expressions simultaneously
• Parallelize operations across CPU cores 

🚀 Full article: https://bit.ly/4hSdN8s 
☕️ Run this code: https://bit.ly/49Io6d0

#Polars #Python #DataScience #DataEngineering

I couldn’t believe it! My book Production-Ready Data Science is featured right on the home page of LeanPub.

Leanpub is a leading platform for publishing and selling self-published technical books, so it’s truly an honor to see my work highlighted there.

The book shares everything I’ve learned about turning data science prototypes into reliable, production-ready systems, from managing dependencies to automating workflows. 

Thank you to everyone who has purchased or shared it. Your support means everything.

Get 58% off using this link (offer valid until November 16): https://bit.ly/43UvgHu 

#Leanpub #DataScience #Python

Convert 8+ document formats to text with the Docling DocumentConverter!

Have you ever needed to pull text from PDFs, Word files, slide decks, or images for a project? Writing a different parser for each format is slow and error-prone.

Docling's DocumentConverter takes care of that by detecting the file type and applying the right parsing method for PDF, DOCX, PPTX, HTML, and images.

Other features of Docling:
• AI-powered image descriptions for searchable diagrams
• Export to pandas DataFrames, JSON, or Markdown
• Structure-preserving output optimized for RAG pipelines
• Built-in chunking strategies for vector databases
• Parallel processing handles large document batches efficiently 

🚀 Article on docling: https://bit.ly/43ANjT8 
☕️ Run this code: https://bit.ly/4og3sow 

#DocumentProcessing #RAG #DataScience #Python

Handle millions of similarity searches in milliseconds with Annoy 🚀

sklearn loads all your item vectors into memory and compares your search vector against every single item in your dataset.

This can take seconds or minutes when you have millions of items.

Annoy (Approximate Nearest Neighbors Oh Yeah), built by Spotify, speeds up similarity search by organizing your vectors into a searchable tree structure.

How it works:
• Pre-builds indexes with `build(n_trees)`, creating multiple trees by recursively splitting your vector space with random hyperplanes
• Traverses tree splits to find the n nearest neighbors using `get_nns_by_item(i, n)`
• Checks only items in the final region instead of scanning everything

As a result, you can query millions of items in milliseconds instead of seconds.

Plus, Annoy is open source! Install it with "pip install annoy". 

☕️ Run this code: https://bit.ly/4833tak 
🔗 Link to the repository: https://lnkd.in/gK9eFKR3 

#MachineLearning #DataScience #Python

Faster, leaner pre-commit hooks with prek (Rust-powered) ⚡

pre-commit is a framework for managing Git hooks that automatically run code quality checks before commits.

However, installing these hook environments (linters, formatters, etc.) can be slow and disk-intensive, especially in CI/CD pipelines where speed matters.

prek is a drop-in replacement for pre-commit that installs hook environments significantly faster while using 50% less disk space.

Built with Rust for maximum performance, prek reduces cache storage from 1.6GB to 810MB (benchmarked on Apache Airflow repository) without changing your workflow.

Key benefits:
• Uses your existing .pre-commit-config.yaml files
• Commands mirror pre-commit syntax (prek install-hooks, prek run)
• Target specific projects/hooks in monorepos
• Install as a single binary with no dependencies 

☕️ Link to the repository: https://lnkd.in/gWaHGrdx 
📬 Want more Python tips? Subscribe to my newsletter at https://bit.ly/4p4WfZX

#Python #DevOps #CI #OpenSource

CodeCut just became a lot more searchable! ☕️

Most developers stick to the same 5-10 Python libraries. Not because others aren't useful, but because finding them is hard. 

CodeCut now organizes 69 tools into eight categories on the homepage: 

• Developer tools
• AI frameworks
• Data processing
• Databases
• Python builtins
• Visualization
• Utilities
• Text processing

Each tool links directly to blogs and code snippets. Browse a new category today. You might find something that changes how you work. 

🍰 Take a look: https://bit.ly/4rJuA1O

Simplify validation rules with Pydantic Field() constraints 🎯

dataclass requires manual validation in __post_init__, separating validation rules from field definitions.

As your data model grows, the __post_init__ method fills with if-else statements, becoming harder to read and maintain.

Pydantic Field() puts constraints directly on field definitions, making your model self-documenting and easier to maintain.

What you can specify with Field():
• Numeric bounds (e.g., age must be >= 0 and <= 150)
• String length (e.g., name must have at least 1 character)
• Regex patterns (e.g., email format validation)
• Default values for optional fields 

🚀 Full article: https://bit.ly/4rILFJr 
☕️ Run this code: https://bit.ly/4pq5m76 

#Python #Pydantic #DataValidation #DataScience

Merge features without memory waste using itertools.chain() 🔗

Standard list merging with extend() or concatenation creates intermediate copies.

This memory overhead becomes significant when processing large lists.

itertools.chain() lazily merges multiple iterables without creating intermediate lists. 

🚀 Full article: https://bit.ly/3JUUP4s 
☕️ Run this code: https://bit.ly/4ozxtjZ 

#Python #DataScience #FeatureEngineering #Itertools

Visualize PySpark DataFrames directly with native plotting API 📊

Visualizing PySpark DataFrames typically requires converting to Pandas first, adding memory overhead and extra processing steps.

PySpark 4.0 adds native Plotly-powered plotting, enabling direct .plot() calls on DataFrames without Pandas conversion. 

🚀 Full article on PySpark 4.0: https://bit.ly/489blG1 
☕️ Run this code: https://bit.ly/484oJuY 

#PySpark #DataScience #Python #ApacheSpark

Transform DataFrames into publication-ready tables with Great Tables 📊

Standard DataFrame output can feel clunky and unfinished. Without clean headers, readable dates, or currency formatting, even great data can look unprofessional.

Great Tables elevates your DataFrames into polished tables built for reports, dashboards, and presentations, all through one chainable interface.

Key features:
• Number formatting: currency, dates, compact notation
• Visual enhancements: mini charts, color gradients, embedded images
• Table structure: headers, subtitles, column control
• Multi-format export: PNG, PDF, HTML

Plus, Great Tables is open source! Install it with "pip install great-tables". 

🚀 Explore all tools: https://bit.ly/4aGHMyv
☕️ Run this code: https://bit.ly/3Mpvtg1

#Python #DataScience #DataVisualization

taplio

Khuyen Tran's Linkedin Analytics

Khuyen Tran

Performance ScoreKhuyen Tran's LinkedIn Performance

What is Khuyen talking about?

Khuyen Tran's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Khuyen Tran's Linkedin Analytics

Khuyen Tran

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Khuyen Tran's LinkedIn Performance

What is Khuyen talking about?

Khuyen Tran's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Khuyen Tran's Linkedin Analytics

Performance ScoreKhuyen Tran's LinkedIn Performance

Khuyen Tran's Best Posts (last 30 days)