Sep 2022 – Jun 2023 · 9 mos · London, England, United Kingdom

Oct 2021 – Aug 2022 · 10 mos

• TrueLayer’s Engineering organization has gone through an amazing period of hypergrowth - from 10 to 100+ in 36 months. I had the opportunity to be there for the entire journey and witness first-hand how our tools and processes had failed to keep up to the challenge.
• I chose to spin up a new team, Engineering Ops, within our Platform group.
• The focus is on the ergonomics and effectiveness of our internal development platform. The mission statement is intentionally broad: boost the impact that every single engineer can have at TrueLayer.
• In less than 5 months we have:
• Overhauled the entire incident management process - per-product oncall rotas, automated setup, all engineers oncall during business hours;
• Given Engineering Managers and Engineering Leads visibility over the operational readiness of their teams (deployment frequency, lead time, failure ratio, MTTR);
• Progressively embedded more and more best practices into our in-house operational tooling, helping teams to fall into the pit of success.
• The team has had a visible impact on the effectiveness of the whole Engineering organization.
• TrueLayer has therefore chosen to double down and invest more: I am now scaling the team to meet our ambitions.

May 2021 – Oct 2021 · 5 mos

• PayDirect found product-market fit - we now had to scale the team and system to handle customer demand, both in terms of volumes and features.
• I partnered with an Engineering Manager to grow the squad from 6 to 20+ engineers - we re-organised into multiple small teams, aligned to the sub-domains required to power our product. I worked closely with Engineering Leads and Senior Engineers to lay down the technical foundations that would support our product growth, execute on key projects (euro currency support) and nurture the next generation of technical leaders.
• My responsibilities extended beyond PayDirect:
• I played a key role in designing and iterating on our first company-wide Engineering growth framework;
• I pitched and drove the adoption of distributed tracing across our entire stack (Rust, C#, Python) to improve our MTTR, alongside other observability improvements (standardized Prometheus metric definitions and practices);
• I worked closely with the CTO, the VP of Engineering and our Engineering Managers to define, evangelize and roll out the company-wide technology roadmap.

Jun 2020 – May 2021 · 11 mos

• Owner of all things tech for our Payouts API product:
• architecture and design;
• delivery and stakeholder management (both internal and external)
• operational support;
• hiring;
• line management of all engineers on the team.
• Payouts API enables merchants to hold funds with TrueLayer and programmatically disburse them as they see fit - we classify it as mission critical software. It is therefore of paramount importance to guarantee a service that is secure, scalable, with well-understood failure modes.
• To support the business requirements we pioneered the introduction of several new architecture patterns, from event sourcing (for auditability and compliance) to event-driven architecture, to ensure decoupling between services and fault tolerance.
• The whole product is built in Rust, a recent addition to TrueLayer's stack. We are particularly keen on Rust's performance, security profile and expressive type system.
• Technology stack: Rust, Postgres, RabbitMQ, Kubernetes, Prometheus, Jaeger

Jan 2020 – Jun 2020 · 5 mos

• The Global Expansion team had one mission in 2020 H1: connect as many European banks as possible.
• In January we had:
• less than 10 engineers;
• few banks available to customers outside of the UK;
• a massive in-progress technical migration.
• My mission was at the crossover between people and technology: I had to scale Global Expansion to meet our ambitious milestones, making changes to ceremonies, processes, team topology as well as evolving the underlying architecture of the system.
• In six months, at the end of June:
• We were live in Spain, Italy, Germany, France and Ireland;
• We had completed the technical migration without any significant incident or misstep, completely transparent to our customers, introducing the concept of dark launches/shadow traffic to TrueLayer for the first time;
• We had grown the team to more than 30 engineers, between FTEs and contractors, across 4 office and 3 timezones. Most members of the original team managed to grow their impact scope and take responsibility for larger parts of the system.

Jun 2019 – Jan 2020 · 7 mos

Mar 2019 – Jun 2019 · 3 mos

• I pivoted into a Backend Engineer role while our customer base was undergoing a phase of hyper-growth: we made sure that our flagship product, Data API, kept satisfying our clients’ needs.
• API requests have grown from thousands to millions per day under my watch.
• Scale came with its own challenges: I was one of the founding members of TrueLayer’s oncall team and I contributed to the introduction of a modern telemetry stack (Prometheus, Grafana, AlertManager) to improve our observability posture.
• Telemetry data allowed us to identify bottlenecks in our user flows, staying ahead of customer expectations when it comes to performance and reliability.
• Technology stack: C# (.NET Core), Redis, Kubernetes, Prometheus, Jaeger, Grafana

Oct 2018 – Mar 2019 · 5 mos

• I joined to enrich TrueLayer’s product with ML-enabled functionality, with a particular focus on TrueLayer's classification engine for bank transactions.
• Over a six months period working elbow to elbow with a fellow Data Scientist we managed to reboot the existing prototype into a fully functional engine - from data ingestion to data labeling, including inference and monitoring.
• Each step of the work was delivered to production, incrementally, working closely with the feedback provided by a few key customers who were heavily invested in seeing the project succeed to leverage its power in their own applications.
• The rebuilt prediction engine significantly increased the accuracy of the system as well as its coverage (% of classified transactions), hitting the desired milestone of 80% coverage across all UK banks.
• Technology stack: Python, Postgres, Redis, Kubernetes

Jun 2020 – Mar 2022 · 1 yr 9 mos

• "Zero To Production In Rust" is an opinionated introduction to backend development using Rust.
• The book can be ordered at https://zero2prod.com

Oct 2017 – Oct 2018 · 1 yr · Londra, Regno Unito

• I built the Data Science department from scratch to support Headstart's machine learning needs: we built a cutting-edge matching system to automatically identify high potential candidates in talent pools of several thousands applicants for big corporate clients.
• Our system allowed recruiters to focus most of their time on the best matches for their openings.
• I was responsible and accountable for the whole lifecycle of machine learning modelling: from prototype and design specification (including customer facing meetings) to production deployment and management.

Jun 2017 – Sep 2017 · 3 mos · Roma, Italia

• Telecommunications / Revenue Assurance:
• Design, development and implementation of a complex ETL pipeline to deal with a consistent amount of customer data (~GBs);
• Exploratory data analysis to support business decisions concerning Offering & Promotions.
• @Tools: Python (Pandas, NumPy, Seaborn), Spark (PySpark interface)

Apr 2017 – Jun 2017 · 2 mos · Trieste, Italia

• I worked as a teaching assistant for the course "Probability and statistics" under the supervision of Dott. Asci and Professor Del Santo.
• My position was paid for by the University of Trieste.

Oct 2016 – Jan 2017 · 3 mos · Trieste, Italia

• I've worked as a teaching assistant for the course "Analisi 3A" (Third analysis course for the students of the B.Sc. in Mathematics) under the supervision of Professor Del Santo.
• My work was paid for by the University of Trieste.