Software Engineer: Applied NLP/ML and Data Systems (Mid-career / Senior)

Theia Insights; Cambridge; Hybrid; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Private health insurance

Pension scheme

Hybrid work in UK (London or Cambridge

Job summary

Theia Insights builds foundational financial intelligence products and serves leading asset managers, hedge funds, and banks. We are seeking an engineer for the Data Products team to own NLP/data pipelines ingesting filings and web content to power TIIC classifications and our thematic factor models.

You will run large-scale NLP inference, ingest market data, and own dataset schemas in collaboration with economists and engineers to ensure reliable production data.

Qualifications

• Strong production Python.
• Datasets in pandas and Parquet/Arrow, plus an analytical engine, e.g. DuckDB, or a warehouse such as Snowflake.
• Orchestrated batch pipelines you've operated, not just written: Dagster or Airflow, S3-based data flows, and a habit of testing outputs for correctness rather than only for exceptions.
• Applied ML in production: embeddings and semantic similarity, clustering, or operationalising models (not necessarily training from scratch).
• AWS fluency and CI/CD discipline.

Responsibilities

• Build and maintain pipelines that classify global public equities across our five-level taxonomy, sector, industry, sub-industry, major theme and micro theme, by extracting information from filings and web content, and assigning thematic exposures based on this information.
• Run large-scale NLP and LLM inference (entity extraction, classification, knowledge graph construction) over company documents, with cost- and throughput-aware batch execution.
• Ingest market data and publish datasets to external distributors.
• Own schema and contract evolution for datasets with real downstream consumers.
• Work with economists and engineers to turn modelling decisions into reliable production data.

Skills

Python

Pandas

DuckDB

Snowflake

Dagster

Airflow

AWS

LLM embeddings

CI/CD

Docker

Tools

Pandas

Parquet/Arrow

DuckDB

Snowflake

Dagster

Airflow

AWS

Docker

SageMaker

Terraform

Job description

About the role

Theia Insights builds foundational financial intelligence products, including industry classification, knowledge graphs and factor risk models, for institutional investors. We serve some of the largest asset managers, hedge funds, index providers and sell-side banks.

As an engineer on the Data Products team you'll own the pipelines that ingest NLP and financial data from public equities around the world to produce the Theia Insights Industry Classification (TIIC) and the datasets behind our Thematic Factor Risk Models (TFM).

The Data Products team owns the data that underpins everything we sell. It's a small, senior group that values correctness and reproducibility over volume, and it sits close to the product leads who shape the methodology. We value durability and good judgement over familiarity with the flashiest tools.

What you'll do

• Build and maintain pipelines that classify global public equities across our five-level taxonomy, sector, industry, sub-industry, major theme and micro theme, by extracting information from filings and web content, and assigning thematic exposures based on this information.


• Run large-scale NLP and LLM inference (entity extraction, classification, knowledge graph construction) over company documents, with cost- and throughput-aware batch execution.


• Ingest market data and publish datasets to external distributors.


• Own schema and contract evolution for datasets with real downstream consumers.


• Work with economists and engineers to turn modelling decisions into reliable production data.

Essential

• Strong production Python.


• Datasets in pandas and Parquet/Arrow, plus an analytical engine, e.g. DuckDB, or a warehouse such as Snowflake.


• Orchestrated batch pipelines you've operated, not just written: Dagster or Airflow, S3-based data flows, and a habit of testing outputs for correctness rather than only for exceptions.


• Applied ML in production: embeddings and semantic similarity, clustering, or operationalising models (not necessarily training from scratch).


• AWS fluency and CI/CD discipline.

Nice to have

• Practical LLM engineering: prompting, batch inference, and cost and throughput trade-offs across providers.


• SageMaker, Bedrock, or comparable managed ML tooling.


• Financial and equities domain knowledge: classification taxonomies, factor models, index construction. Valuable but learnable.


• Infrastructure as code (AWS CDK or Terraform) and Docker.

Experience we're looking for

We care more about what you've owned than years on a CV. If you've built a pipeline that runs on a schedule against real volume, and you were the person who got paged when it broke, you're in scope. More senior candidates will typically have made the cost and throughput trade-offs, and set the standard for how a team tests data quality.

• Competitive salary plus EMI share options


• 25 days holiday + bank holidays


• Private health insurance, pension


• Hybrid working in UK from London or Cambridge