Like Runwise, Pulley, Callbox and Sifted.
Data & Analytics Technologies We Work With
Layer by Layer Modern Data Stack
A modern data stack refers to a set of cloud services that move data from the collection point to the decision-making point, including collection, storage, processing, and visualisation options. We are using your existing stack while suggesting new technology only when it is worth it.
Technology We Use
Every mentioned technology is used in one of the projects specified in our portfolio.
“Used in X number of projects” means “the projects in the portfolio are numbered”. Companies are named as they are listed in our portfolio.
-
Cloud and Compute
Serverless jobs and containers that execute your pipelines without need to manage servers.
AWSNew consulting page is coming soon Lambda and ECS for pipeline jobs and DynamoDB for high-speed lookups.USED IN 5 DELIVERED PROJECTSData engineeringDockerServices that can be reused multiple times and act the same in production and development.USED IN 5 DELIVERED PROJECTSData engineeringIncluding Chili Piper, Agora, and Banyan Technology
-
Analysis & Machine Learning
From the clean tables to forecasts and predictions.
PythonThe main language we use to develop pipelines, analyze data, and create models.Used in all of the 21 published projectsPredictive analyticsEvery project from our portfolio
Pandas & SQLAlchemyPreparing and modeling data in code that is readable and testable.Used in 6 published projectsData analyticsAmong them: Sifted, Pulley, FourKites, Census
-
Warehouses and Databases
Where data is kept and prepared for analysis.
SnowflakeConsulting page is up next Cloud-based warehouse for heavy-duty analytics and ingestion.USED IN 4 DELIVERED PROJECTSData engineeringDatasembly, Census, 4Kites and Aera Technology
Google BigQueryConsulting page is up next Serverless warehouse for event and marketing information.USED IN 3 DELIVERED PROJECTSDigible, Aera Technology and Root Insurance
PostgreSQLThe database providing reliable storage for most of the data applications we deal with.USED IN 7 DELIVERED PROJECTSData engineeringIncluding Runwise, Sifted, Chili Piper and Agora
RedisIn-memory queues and caching for real-time features.USED IN 6 DELIVERED PROJECTSData engineeringIncluding ShelterZoom, Sendoso and Attentive
-
Pipelines, Streaming and Orchestration
Data transmission by schedule or in real time with retries and alerts.
Apache AirflowPipelines that can be scheduled and monitored instead of depending on erratic cron jobs.USE OF 3 PUBLISHED PROJECTSData engineeringCensus, Aera Technology, Root Insurance
Apache KafkaOur consulting page will be launched soon. Stream processing for events that need to take place without waiting for night’s batch processing.USE OF 2 PUBLISHED PROJECTSDatasembly, Root Insurance
Scrapy & BeautifulSoupCollecting data from the Web in such a way as to be sure that it won’t fail silently.USE OF 4 PUBLISHED PROJECTSData miningAmong them: Digible, Callbox, Roofstock
FastAPI & Django RESTAPI to provide your data and models to other systems.USE OF 9 PUBLISHED PROJECTSData engineeringAmong them Chili Piper, Innovaccer, Banyan Technology
Also in Our Toolkit
In this category, we provide tools that our team is familiar with, even if we have not had the opportunity to work on a project with a client yet. Ask us during a call to find out the kind of tools we use.
- Power BI
- Tableau
- Excel
- Microsoft Azure
- Microsoft Fabric
- Databricks
- dbt
- Google Cloud (beyond BigQuery)
- TensorFlow
- PyTorch
How We Choose The Right Platform For You
Most projects start at a stack you already have.
-
Volume and speed
Nightly batch, hourly, or real-time streaming
-
Cost structure
Pricing for compute, storage and data transfer.
-
Who maintains it
Your team’s capabilities after we finish the process.
-
What you’re currently using
Licenses, clouds, and tools that your team is familiar with. (Button: Send us your stack)
-
Governance
It refers to safeguarding access, auditing, as well as locations for data storage.
We have no vendor affiliation and earn no referral fee from insurance companies, which means that our recommendations are purely based on your requirements, not on sales targets.