Everything Beady does starts with the pipelines that pull from 100,000+ public sources — company registries, court and enforcement records, sanctions and watchlists, news and social. You would own the part that keeps them flowing, complete and current.

What you would do

• Build and maintain collectors for registries, court systems, regulators, sanctions lists and media
• Keep sources alive when they change format, rate-limit, or simply go down
• Design storage and indexing for hundreds of millions of entities and their documents
• Make every record traceable back to the page it came from

What we look for

• Strong Python (or Go / Java) and SQL, comfortable with queues, schedulers and distributed workers
• Real experience with large-scale crawling or ETL and with messy, inconsistent data
• A bias for correctness — a record we miss is not a bug, it is a compliance failure

Send your CV and two lines about why this role to hello@beady.ai.

Most of what a name search returns is noise. Our job is to remove around 95% of it without ever losing the one hit that matters. This role owns that trade-off.

What you would do

• Match entities across spellings, transliterations and aliases in dozens of languages
• Train and evaluate models that tell a real hit from a coincidence — same name, different person
• Classify adverse media by relevance, severity and source credibility
• Own the metrics: false positives are expensive, false negatives are unacceptable

What we look for

• Production NLP and ML, not only notebooks
• Experience with entity resolution, record linkage, NER or search relevance
• Multilingual text and non-Latin scripts are a strong plus

Send your CV and two lines about why this role to hello@beady.ai.

Someone has to know where the data actually lives, how far it can be trusted and what a finding really means. That is this role.

What you would do

• Map and prioritise new public sources: registries, courts, regulators, law enforcement, media
• Define what a confirmed hit looks like for each source — and where that source stops being reliable
• Validate system output against manual research and hand the gaps back to the ML team
• Shape the methodology behind our reports so they hold up in an audit

What we look for

• Hands-on OSINT, investigative, due diligence or AML research experience
• Comfortable reading primary sources in more than one language
• Precise and sceptical, and able to explain why a finding is — or is not — real

Send your CV and two lines about why this role to hello@beady.ai.

We build for compliance officers, so we want one on the inside: someone who has run the checks, signed off the files and knows exactly where the process hurts.

What you would do

• Turn AML, KYC, KYB and sanctions requirements into product decisions
• Work with customers’ compliance teams on how they screen, review and record decisions
• Shape reports and audit trails so they answer the questions a regulator actually asks
• Back the team up on demos and on the hard customer questions

What we look for

• Experience in a compliance, AML or due diligence function at a bank, fintech, fund or law firm
• Working knowledge of the major sanctions regimes and of screening practice
• Able to write clearly for regulators and for engineers

Send your CV and two lines about why this role to hello@beady.ai.

Yes — four roles are open right now and listed above. If none of them is quite you, send your CV anyway: we hire ahead of need, read every one and reply.

Engineering, ML and NLP, data and OSINT research, compliance domain expertise, and sales and partnerships.

No. Curiosity about how risk hides in public data matters more, and we’ll teach you the rest. For analyst and domain roles, hands-on KYC, KYB or AML experience is a strong plus.

Four steps: your CV, a 30-minute intro call, a short practical session on a real problem, and an offer — usually within two to three weeks end to end.

Working English for everything customer- and product-facing.