Consilient brings federated learning to financial crime problems.I am driving, defining and coding the technical build-out of the core federated technology platform, especially the role of privacy enhancing technologies including differential privacy and confidential computing and using my extensive experience with inhouse analytics platforms to design for deployability to banking clients which I also in a product success role.I develop our federation system using neural networks, writing fully deployable pipelines and a federated learning platform based on Azure using typescript, react, python/pytorch, spark and dbt.

I decided to leave HSBC following the launch of their next-gen AML capability for a career break whilst starting getting involved in various startup opportunities.CTO/full-stack developer for a bootstrapped, novel spaced-memorization app for education targeting android, iOS and web as a venture with examstudyexpert.com.I built the app from the ground up and ran it through trials of up to 100 simultaneous users. Ongoing effort to scale-out and monetize.

Building on the platforms I'd put together, I went on to product own feature engineering and data science optimisation pods, comprising of approximately 30 developers for a supervised learning based replacement for anti-money laundering transaction monitoring.I was proud to deliver a coherent end to end data science pipeline into a dual running production in a very difficult space, tackling governance, technical, data science and cybersecurity problems.Left shortly before go-live for a paternity career break.

I worked with big tech organisation to build a collaboration environment to work with bank developers, architecting and engaging with teams across the bank to get sign-off to implement and then move to implementation on GCP.I led a development team building a first-in-bank collaboration environment allowing tech organisation on GCP, with a focus on cybersecurity and platform security (whilst allowing productivity).

Promoted from my role to lead analytical tooling efforts, managing the tooling on two hadoop clusters.I led a small team providing tooling to several hundred analysts building transaction monitoring systems and helped them work more effectively with the scale of the problems they were facing, including building a secure on-cloud and on prem development environment. This supported various AI, ML and more traditional rules based initiatives across financial crime.* Acted as an advocate for use of newer software on a "secure analytics development platform".* Encouraged standards, helped to resolve critical issues and ran weekly best practice workshops as an advocate for best practices eg. version control.* I substantially reduced SAS usage onto open source python, reducing spend by 75%.* Led architecture and roll-out of first "on-cloud" analytics environment in the bank (GCP).* Product owned rollout of Qliksense platform.* Product owned Counterparty banking transaction monitoring system implementation.

Leading on from my effort putting together a basic data model and handing off the modelling role to a global rollout, I started a role on getting the bank leveraging the technologies which they'd just invested, helping the FIU (Financial Intelligence Unit) leverage its considerable investment into its hadoop platform.Key Projects:Proposed, obtained funding, designed and led implementation of HBASE (NoSQL) fast transaction lookup service, reducing time for analysts to obtain transaction data to seconds from up to 30 minutes during.Led evaluation of various vendor products, onboarded Datameer to assist non-technical users with their use of HadoopLed implementation of and rollout of Apache Spark for Data Science teamParticipated in "rapid response" to Panama Papers, including setting up an elasticsearch engine across the bank. I'm particularly proud that I managed to obtain a meeting with a key journalist at an industry event which gave me a pre-release insight into the data structure

I led small team modelling a common data schema across the bank's three largest jurisdictions and three different core banking platforms.I led several iterations across different developer teams, prototyping development myself.This effort evolved into a much larger data modelling effort which continued to this data, and this the output of this rapid prototyping was used for almost 5 years across many financial crime projects and served the basis of a much larger common data effort.

- Using tableau and SAS skills to identify and detect fraudulent activity and market manipulation in a variety of market datasets and for a number of clients across a range of different industries. - Client facing, investigations into complex and difficult to resolve problems using innovative techniques. - Involved in a number of successful bids to organisations in disparate sectors. Technical Skills: Tableau, SAS, Python (Pandas), Django, Unsupervised Learning Random Forest, RoC, Statistical and Situational Analysis

- Solely responsible for implemention of a customer change request whilst on a direct customer facing role in a Government role, working closely with clients- Tier one investment bank, unauthorised trading platform - wrote scoring and assessment framework working directly with clients in a pre-sales setting- Investment bank, unauthorised trading platform - consolidated existing skills as lead full-time analyst on the project- Contributed to the development of Detica's core unauthorised trading offering, in particular analytics and reporting functionality.- interest in the both the technical challenge of unauthorised trading monitoring and the business cases behind them- HSBC bank - counterparty fraud solution, scoring and analytics lead, working with business to improve performance of automatic case assessment on millions of entities and terabytes of data. Led development of scoring an analytics framework and rapidly became system SME. Led analytic and technical workstreams and presented results to business users, giving opinions and improving pattern detection, holistic scorecard development and scenario identification and database development and design. Wrote complex network scores in the AML, Sanctions, PEP and watchlist spaces for well over a year. - Came up with a testing strategy for a complex set of algorithms in an inherently difficult-to-test language (SAS) and developed a testing framework in Java, implemented a Jenkins deployment and wrote a simple ANTLR based SAS language tester for verification. - Ran a Globally deployed system with a 1 week end-to-end batch time, providing support and acting as primary troubleshooter for a complex dynamic system, some of which required complex patches which I would architect and implement.Technical skills: Linux, Oracle, Python, SAS, Agile Development, Technical Assurance, Bash scripting, ANTLR, Java, J2EE webapps, Birt, Selenium, Maven, Mercurial

Took six months out from Degree at Durham.- Worked on the Detica NetReveal solution implementation team, primarily as a DataServer (SAS) developer- Activities included analyzing data running into the hundreds of gigabytes for a major UK industry body, including development the development and deployment of bespoke pieces of automated analytics with a requirement for reliability and accuracy- Worked closely with clients to develop their needs- Also worked on the development of bids to international law enforcement.At the conclusion of my internship I was offered a permanent position.

- Created recommendations for changes to the process of monitoring the negative financial impacts of engineering changes and presented ideas to senior stakeholders at conclusion of internship and scheduled these to be implemented.- Developed and handed over change control tracking software in VBA/excel for the use of change control team to visually track the progress of engineering change requests.

- Developed layout and added additional content to university level one physics website- Worked with very little supervision and a very open brief- Layout is still in use today and has been extended significantly by others