Lokesh Arora

About

Engineering systems that scale.

I am a senior backend developer with fifteen years of experience in Java and adjacent backend technologies, building scalable, high-performance applications for financial technology platforms.

Most of that time has been at BlackRock in London, where I work as a Vice President and development lead for the Aladdin Channels and Genie teams. Aladdin Channels builds the integrated digital client engagement experience behind the Aladdin client sites; Genie is the desktop application launcher Aladdin clients open to start their working day. Before that I spent seven years on BlackRock's public-facing retail websites and digital tools, including a product screener that helps investors find the right products by surfacing the information that actually matters.

Earlier roles took me through credit risk systems at Royal Bank of Scotland, order management and CPQ integration at Accenture, and derivative pricing data services at Infosys — consistently the systems where correctness, confidentiality and uptime are not negotiable.

The work I care about is designing and optimising microservice architectures, cloud deployments and secure APIs, and leading the people who build them. Lately that curiosity has extended into AI engineering — treating models as another distributed system component with latency, failure modes and cost to design around.

Experience

Fifteen years, four organisations, one discipline.

From derivative pricing services to credit risk platforms to the technology institutional investors run on — a steady progression from developer to development lead.

2010

Infosys

System Engineer

2013

Accenture

Software Engineer Analyst

2014

RBS

Associate

2015

BlackRock

Vice President

Present

Dev Lead

Aladdin Channels & Genie

Vice President

BlackRock

Aladdin Channels & Genie

06/2015 — Present

London

Development lead for the Aladdin Channels and Genie teams. Aladdin Channels is responsible for the integrated digital client engagement experience across the Aladdin client sites; Genie is the application launcher Aladdin clients use to start their business day.
Role detail
- Working as a development lead for the Aladdin Channels and Genie team.
- Aladdin offers clients the same technology and intellectual capital BlackRock uses to manage investments day to day.
- Since joining BlackRock, worked across a range of Aladdin and non-Aladdin applications.
- Previously responsible for overseeing development and optimisation of web experiences and digital tools for BlackRock's public-facing websites, as part of the Digital Marketing team.
- Developed a product screener tool that helps knowledgeable investors find the right products more effectively by surfacing the essential product information.
- Gathering requirements with clients and product owners, managing BAU development, month-end releases and production support.
JavaSpring BootMicroservices REST APIsKubernetesAzure Teamsite (CMS)GrafanaADO
Associate

Royal Bank of Scotland

Project · PRiSM

03/2014 — 06/2015

Gurgaon, India

Senior developer on multiple credit risk applications within the PRiSM suite, used to manage corporate loans across the organisation.
Role detail
- Developed and maintained multiple credit-risk applications under the PRiSM suite.
- PRiSM comprises Credit Submission, RDC, EMR and TUNA, and is used for managing corporate loans.
- Credit Submission is an RBS-owned limit and workflow management tool within the suite.
- The system holds confidential customer information and is strictly for internal use.
JavaSpringWeb Services MySQLWorkflow management
Software Engineer Analyst

Accenture Pvt. Ltd

Project · Comcast CPQ Integration

02/2013 — 03/2014

Pune, India

Built and maintained the CPQ (Configure, Price, Quote) module for the IBM Sterling product, synchronising generated quotes with the Comcast environment.
Role detail
- Created and maintained the CPQ module for the IBM Sterling product.
- The platform lets users create quotes and opportunities based on business requirements.
- Generated quotes are synchronised with the Comcast environment using SOAP web services.
- Facilitates the flow of business data that converts opportunities into actual transactions.
- Primarily used for order management, ensuring efficient and streamlined processing.
JavaIBM SterlingSOAP Web Services Order Management
System Engineer

Infosys Ltd

Project · Trade Capture Data Services

07/2010 — 02/2013

Pune, India

Worked on the curve values needed to price derivative products, exposed to downstream systems through web services.
Role detail
- Determined the curve value needed to price derivative products in derivative trading applications.
- Built the GUI used to manage the various factors that determine the curve, and to report on the data.
- Designed and implemented parsers for flat files.
- Developed RESTful web services through which the calculated curve is fetched by other systems.
- Developed Autosys jobs running on schedule to perform different tasks.
JavaRESTWeb Services AutosysFlat-file parsers

Selected work · BlackRock

The systems I lead.

Three bodies of work inside BlackRock, described at the level the public record allows.

01

Aladdin Channels

07/2022 — Present

Leading development and maintenance of the Aladdin client sites (Aladdin Hub). Building reusable Teamsite CMS components for the ACS sites and maintaining the financial tools institutional investors use, alongside BAU development, month-end releases, and requirements gathering with clients and product owners.

JavaMicroservicesTeamsite (CMS)REST APIs

02

Aladdin Genie

01/2024 — Present

Leading development and support for Genie, the Aladdin financial desktop application launcher. Responsible for design through to deployment on client environments, managing BAU development and month-end releases, and providing critical on-call support for production issues.

JavaSpring BootProduction supportOn-call

03

Core Sites & Architecture

06/2015 — 06/2022

Developing and maintaining BlackRock's retail websites across the globe. Key contributor on product pages, the product screener and various financial tools, plus Teamsite CMS component development and monthly BAU releases.

JavaTeamsite (CMS)Web ServicesComponent design

In progress

Building what's next.

Independent work outside the day job, where the distributed-systems habits meet newer tooling. These are in development — listed here because the direction matters more than the demo.

Personal project

AI Knowledge Platform

A retrieval-backed knowledge service: document ingestion, embedding storage and a query API, designed so the model sits behind the same latency, failure and cost boundaries as any other downstream dependency.

JavaPythonREST APIsRedis

Coming soon

Personal project

Intelligent Portfolio Platform

An investment tooling sandbox: portfolio modelling services, a clean domain API, and an interface for exploring holdings — a personal take on the product problems I have spent a career near.

Spring BootGraphQLMySQLDocker

Coming soon

Personal project

Distributed Systems Lab

A running notebook of implementations: message-driven services, idempotent consumers, caching strategies and failure injection — built to test the trade-offs rather than to read about them.

GoKubernetesGrafanaLinux

Coming soon

Architecture

How I think about systems.

Three reference designs, written as personal architecture exercises rather than descriptions of any employer's production systems. Open one to see the shape and the trade-offs.

01

Scalable Microservices Platform

Architecture case study · personal exercise

Open

Scalability: Services are stateless and horizontally scaled behind the gateway, so capacity is a replica count rather than a redesign. State is pushed to the datastore and a cache tier that absorbs read-heavy traffic.
Reliability: Timeouts, retries with backoff and circuit breakers at every network hop; health and readiness probes so orchestration removes a sick instance before users notice it.
Security: Authentication and rate limiting terminate at the edge; services trust a verified identity rather than re-implementing auth. Secrets stay outside the image, and traffic between services is encrypted.
Observability: One correlation ID follows a request across every hop. Dashboards are built around latency percentiles and error rate, not averages — averages hide exactly the users who are suffering.

Trade-off. Splitting a system into services buys independent deploys and scaling, and charges for it in network failure modes, distributed debugging and operational surface area. Worth it when teams need to move independently; expensive when they don't.

02

Event-Driven Backend

Architecture case study · personal exercise

Open

Scalability: Partitioned topics let consumer groups scale with the workload while preserving ordering where it actually matters — per key, not globally. Slow consumers create lag, not outages upstream.
Reliability: Consumers are idempotent, so at-least-once delivery is safe. Poison messages land in a dead-letter topic with enough context to replay them once the bug is fixed, instead of blocking the partition.
Security: Topic-level authorisation keeps each service to the streams it needs. Events carry only the fields consumers are entitled to; sensitive payloads are referenced, not copied into the log.
Observability: Consumer lag is the first-class signal — it is the earliest honest measure that the system is falling behind. Alert on lag trend, not on a single breach.

Trade-off. Asynchrony decouples producers from consumers and makes the system resilient to downstream slowness, at the cost of eventual consistency and the need to explain to users why something they just did is not visible yet.

03

AI Application Architecture

Architecture case study · personal exercise

Open

Scalability: The model call is the slow, expensive hop, so it is isolated behind its own service with queueing, concurrency limits and response caching. Retrieval scales separately from generation.
Reliability: Every model dependency gets a timeout, a retry budget and a defined fallback — a smaller model, a cached answer, or an honest "unavailable". A degraded answer beats a hanging request.
Security: Inputs and retrieved context are treated as untrusted. Retrieval respects the caller's entitlements, so the model can only ground an answer in documents the user was already allowed to read.
Observability: Latency and error rate are not enough: quality needs its own evaluation set, and cost per request needs the same dashboard treatment as any other resource.

Trade-off. Retrieval-grounded generation makes answers traceable to sources and far easier to trust, and adds an indexing pipeline, a second storage system and a freshness problem to own.

Technical skills

The toolkit.

Technologies I work with in practice, grouped by what they are for.

Backend

JavaSpringSpring Boot Spring SecurityRESTful APIsGraphQL gRPCMicroservicesWeb Services

Cloud & platform

AzureKubernetesDocker CI/CD PipelinesLinux

Data & performance

MySQLRedisGrafana

Engineering & delivery

GitJUnitMaven ADOTeamsite (CMS)

Also working with

PythonGo

Languages

English — native / bilingualHindi — native / bilingual

Education

B.Tech, Computer Science

Graphic Era Institute of Technology · Dehradun, India

06/2006 — 05/2010

Contact

Let's build something meaningful.

For engineering leadership, backend architecture, AI engineering and technology opportunities — London or remote. The quickest route is email; I read everything.

Email arora.lokesh@hotmail.com Curriculum vitae View or download PDF

Based in London, United Kingdom GMT / BST