Lead Data Scientist - Argentina

Key Skills

AWS, Azure, Continuous Integration & Continuous Delivery - CI/CD, Data science techniques, Effective communication skills, Google Cloud Platform (GCP), Leadership Skill, Machine learning techniques, Python Programming, PyTorch, Scikit-learn, SciPy, Signal processing, TensorFlow

Job Description

About This Role

As a Generalist Lead Data Scientist at Very, you will work with our Software, Hardware, and Product Design teams to build full-service solutions for our clients. We focus on building end-to-end hardware and software solutions that meet our client's business needs, and reliable machine learning systems are often a part of our offering. An ideal candidate will display technical expertise in machine learning and data science, as well as strong communication skills, and the ability to present ideas and results to audiences ranging in technical depth. Candidates should also have experience translating business problems into analytical solutions, working in interdisciplinary teams, and building ML models for production systems.

What You’ll Be Working On

At Very, there is a never-ending supply of variety to the types of projects we work on. However, it is critical to note that almost all of these projects are production systems. As such, the only consistent research component of this position will revolve around establishing a pattern of delivery that allows the team to implement full-scale applications leveraging machine learning in a fast, predictable manner. 

You’ll spend 80-90% of your time working on products or platforms for one of our clients, and the other 10-20% of your time will be spent improving Very’s Data Science Practice. This will involve:

- Working with other data scientists to continuously improve our delivery process for data science applications.

- Working with our marketing team to generate high-quality content (blog posts, conference presentations)

- Working with our sales team to close deals and build meaningful, well-scoped proposals for potential clients.

As a lead data scientist, you will be expected to not only understand the following analytical approaches but also be able to guide others in their implementation:

- Predictive modeling and/or anomaly detection on multivariate time series data

- State classification and prediction for geospatial time series data

- Regression and classification using a variety of deep learning and ensemble tree-based methods

- Clustering/segmentation

- Dimensionality reduction or latent space representation

- Classical statistical analysis and signal processing

Our Current Tooling

Our data science contracts typically involve building a greenfield API or greenfield product from the ground up. In the context of the data science and machine learning pipelines, we typically leverage:

- Git, GitHub, CircleCI or GitHub Actions (CI/CD), pytest (TDD)

- The standard SciPy Stack (Numpy, SciPy, Pandas, Scikit-Learn, Matplotlib)

- SQL/Postgres

- Docker

- Jupyter notebooks for prototyping

- Cloud architecture and resources for production systems. eg)

-- AWS: Lambda, ECR/ECS, RDS, API Gateway, Batch, Sagemaker

-- Azure: Functions, Container Registry/Instances, SQL Database [for PostgreSQL/MySQL], API Management, Machine Learning

-- 3rd party: TimescaleDB

- Serverless or Terraform for Infrastructure-as-Code

- MLflow

- PyTorch

On our full-service builds, we often reach for the following tools. Experience with them is not required, but any familiarity with these tools is a plus. Our build teams operate with a very high degree of collaboration, so you will definitely have run-ins with these stacks throughout your time here:

- Python web development frameworks (Django, Flask)

- React & React Native

- Swift & Objective C

- Ruby on Rails

As an IoT technology company our data science pipelines include “Things”. This will require you to build pipelines and deploy analytical models to hardware on the edge in addition to the cloud. This requires a deep collaboration with the design, software and hardware teams in the following environments:

- Elixir, Phoenix, and Nerves

- Embedded C and other lower level languages such as Rust

- CI/CD including hardware and end-to-end testing and verification

- Development Single-Board Computers such as RPi

At Very we have two types of lead positions; a specialist and generalist. Both positions go beyond the mastery of the core Data Science competencies to include a leadership position. The specialist lead focuses in great depth in one or more areas of Data Science and spreads that mastery across Very. The generalist lead takes a more multi-disciplinary approach and leads multi-disciplinary projects from a technical perspective.

Responsibilities

-- Work with stakeholders (clients, sales, engineers & designers) to define Statements Of Work (SOW)

-- Communicate how data science can deliver value to the client

-- Estimate quantity of work required to unlock this value

-- Identify related assumptions, risks and dependencies 

- Take ownership of the data science components and related systems to ensure project success

-- Architect, build and deploy reliable end-to-end data and ML pipelines into production

-- Ensure the highest level of testing across the full data pipeline and operating envelope

-- Execute and document all algorithm verification testing for certification

-- Build and nurture strong relationships with clients, understand their perspective and walk them through the data science value chain

-- Facilitate complex conversations to achieve alignment to drive positive outcomes

- Continue to expand and evolve the Data Science (DS) practice at Very

-- Provide technical guidance to DS and non-DS team members

-- Pairs with mid team members to develop their skills and deliver on projects 

-- Continually learn, share and refine your DS skills and knowledge

-- Establish and enact DataOps and MLOps best practices 

- Take on the responsibility of Technical Lead on complex multi-disciplinary projects

- - Build reliable product roadmaps with technical implementation strategies

- - Monitor and optimize the technical implementation and coordination of the project

- - Identify and mitigate risks and seek assistance when required

- As a client services organization, travel may be required up to 10% of the time.

Qualifications

Required

- Master’s degree in Data Science related field

- 5+ years of related experience

-- Deployed statistical, ML or other analytical models to production

-- Lead teams with hardware and software engineers

-- Performed real-time signal processing or lead teams that did so

- Strong written and spoken communication skills in English

- Proficient in Agile development

- Mastery in analytical framing

-- Breaking down solutions into “thin vertical slices” of work 

-- Guiding interdisciplinary team to successfully estimate and execute these slices

- Deep understanding of all four types analytics (Descriptive, Diagnostic, Predictive & Prescriptive)

-- Translating desired client outcomes to the appropriate type

-- Understanding the data requirements and techniques to achieve each type

- Experience using a cloud computing platform like AWS, GCP, or Azure

- Expert level Python development skills related to Data Science

-- Object-Oriented Programming

-- Automated testing, code coverage, model building & evaluation

-- SciPy Stack, Scikit-learn, Tensorflow or PyTorch

-- GitHub CI/CD best practices 

- Experience developing in, compiling and deploying low level software; C++, Embedded C or Rust preferred

- Proficient developing in Linux including and light administration

Nice-to-have

- 7+ years of related experience including with connected devices

- Proficient in embedded real-time signal processing

- AWS Professional level certification

- Familiarity with Elixir, Phoenix, and Nerves

- Hands on experience with Single-Board Computers such as RPi

Contract

Your contract with Very will be as an independent contractor for an indefinite period.

This type of contract carries some additional responsibilities for you: You'll have to save some money to take care of taxes, and you'll have to pay for public social security (Health and Pension) out of pocket.

That being said, we're confident that our competitive compensation and benefits package (see below) will still mean that you are happy with your take-home amount at the end of each month.