Siddharth Agnihotri

Leading two software development teams of 10+ resources. Own multiple distributed systems responsible for processing various types of Market Data (Prices, FX, Rates, Corp Actions, Indices/ETFs Compositions and more) from various internal/external upstream (Reuters, Bloomberg, ULTUMUS, SOLA etc. ) and publish into different backend databases DMDS (OODB), Sophis (Oracle DB). This data is used globally by Front Office and Middle Office teams for Trading, PnL calculations, post trade feeds generation etc.

Systems are developed in JAVA, C++, C#, .NET etc. and use Object Oriented (DMDS, MongoDB) and Oracle Databases for storage. Strategic systems are containerized (Docker) and run on internal Kubernetes platform. Integrated with EQD Hosting Platform services for logging, monitoring and authentication/authorization.

Roles and responsibilities:

• Regular connects with business partners to maintain continuously moving projects pipeline, roadmaps
• Prioritization and strategic future state planning
• End to end ownership of technical deliveries and maintenance across all phases of project lifecycle such as requirement gathering, system design and development, Go Live, L2 & L3 support
• Team's growth and expansion in terms of strength, skills, diversity and maturity

Key Technical Achievements:

Equities Corporate Actions Strategy (WIP):

• Developed in JAVA with DevOps integrated. Built following TDD approach.
• A comprehensive strategic solution cross Equities – Completely automated real time processing of Corporate Actions events. Event converted into bookable trades and book into downstream systems within the SLA
• Use cases implemented around ~50 types of Corporate Actions (Stock Split, Cash Dividends etc.)
• Business Benefits: Removes need of ~40K complex manual amends by Ops Business annually. Results in huge reduction in Operational Risk and ensures process transparency to business

Equity Derivatives Global Price Fixings Solution:

• An easy to scale system which is developed in JAVA and integrated with DevOps
• Strategic solution for automated Fixings of Official EOD Prices (Open, High, Low, Close, EDSP etc.) for Stocks & Indices
• Identifies required Fixings through Quants libraries, fetches EOD Fixings from Bloomberg & Refinitiv and flows into Downstream systems
• Business Benefits: Accurate and timely fixings. ~60 hours of manual efforts savings for Ops Business. Uses trusted and centralized price source saving $1.75 Million annually through caching and licensing cost. Operational risk reduction. Allows business to scale w/o adding to manual efforts

Cash Equity Trade Aggregation:

• Aggregation of Cash Equity Trades in Sophis, position and book keeping system
• Achieved 50% reduction in DB Volume
• Drove this delivery coordinating/collaborating across 20+ team in Equity
• Savings of ~9K Hrs of daily job runtime resulting in Grid, CPU & infra cost savings
• Business Benefits: No SLA breaches for many critical feeds delivered to Regulatory, Finance, PnL, Risk & Control teams

Sophis Capacity Metrics on Grafana:

• Developed Grafana Dashboard for visualization of various KPIs to monitor Sophis performance
• Key data points taken from Sophis DB, calculated KPIs stored into MongoDB and rendered on Grafana
• Business Benefits: Dashboard is widely used by various technology and business teams for real time monitoring of various Finance PC, OPS, regulatory, Risk job runs, as well as to predict future performance of the system

Led a software development team of ~10 resources. Owned end to end development life cycle of Online Fraud Portfolio which comprises of 14 applications used by Fraud Risk Management team for Credit Card Fraud Identification and Fraud Servicing. Using these applications users view selective raw/enriched data and take appropriate actions such as application of DAC Rules, Status flags etc. which apply on potentially compromised accounts in real time. Data provided by these applications also help analysts/business users to formulate their fraud identification/servicing strategies. Applications were built in JAVA and use Teradata and Sybase Databases. Applications had web based front end and were used globally.

Key Technical Systems:

CACTUS (Complete Application for Compromise and Testing based on User Specifications):

• Used to identify potentially compromised merchants.
• Identifies CPP (Common Point of Purchase) and CPT (Common Point of Testing).
• 400+ modules, which users can add/modify dynamically, run daily to identify potentially compromised merchants.
• Monitors risk activity on potentially compromised accounts and generate several types of MIS reports.

AUTOCARD:

• Used for application of DAC Rules and Status Codes on potentially compromised accounts.
• Receives compromised accounts details from upstream applications (CACTUS, RSMS) and users can manually feed the details as well.
• Used to process card replacement requests in case of renewal, expiration, worn out and fraud.

SCREF (Suspect Claim Referrals):

• Used for providing analysts with the fraudulent transactions (Referrals) details analyzing which it is decided who should be re-billed (Card member or Merchant) and whether partially or fully.
• Receives fraudulent transaction details directly from card members or based on the backend processing (Scanners). Risk score calculation is done using a Machine Learning Model (GBM Model).

enPart of Equity Linked Technology team – Worked on multiple applications which deal with Equity Derivatives Data and are used globally by Front Office and Middle Office users. My applications spanned across life cycle management of corporate actions, real time monitoring of barrier options, historical representation of firm wide product data/market data changes, P&L swings calculations through on-the-fly pricing and clean reference data maintenance in RAM (Risk Analysis and Management System).

Projects, Roles and Responsibilities:

Worked as a developer on following systems and my role extends to get the business requirements from the scratch, convert them into technical requirements, implement them, manage production releases and provide L2, L3 support.

Barrier Options Monitoring Service (BMS):

• JAVA application used by Front office and Middle office users globally.
• Used for monitoring live movement of barrier options (intraday and close) based on price ticks from sources such as Reuters, Bloomberg, PARC and Synthetic.
• Users can perform actions on the barrier states (accept, reject, discard) via Application UI.

Quartz History Viewer:

• Developed using Quartz Framework (Proprietary to Bank of America) and Python.
• Used globally by Middle Office users to verify P&L impacts and to investigate on the breakages.
• Leverages bi-temporal property of the Sandra DB (OODB) to view historical changes to the product data.
• On-the-fly pricing before and after product data changes to calculate PnL swings.

Hydrogen (Reference Data Management):

• Developed in JAVA and used by Front Office, Middle Office users globally for maintaining clean Equity Derivative product data (Create, Modify and Search) in RAM.
• Keeps RAM in sync with upstream golden reference data sources PME and ETD.
• Supports auto update/auto create functionalities for specified instrument/model combinations.

Dubnium (Risk Reporting Framework):

• Developed in JAVA and generates end of day risk reports which are used by traders and risk managers to assess exposure to risk by calculating Var (Value at Risk).
• Reports are generated for various hypothetical scenarios (Market parameter shift) decided by RMs.
• Reports contain Greekvals (deltaval, gammaval, thetaval, Vegaval, partialVegaByMaturityVal etc.) and positionvals calculated under each scenario.
• Generates thousands of reports for all the folios with live positions globally.