Reproducibility
Versioned inputs, explicit assumptions, and deterministic workflows make research easier to audit and extend.
HELLO, I’M
Filipe Soares
Quantitative Developer
I transform market hypotheses into institutional-grade trading systems—from rigorous research and validation to production deployment.
Research systemsRisk analyticsTrading automation
01 / ABOUT
Engineering clarity
into quantitative work.
My work is centered on quantitative software engineering: research and backtesting infrastructure, market-data systems, risk analytics, and trading automation. The goal is software that can be reviewed, reproduced, and maintained—not a black box.
Every engagement starts by making assumptions explicit, defining validation criteria, and choosing the smallest architecture that can support the decision at hand.
Data integrity
Time alignment, point-in-time availability, and quality checks are treated as system requirements.
Clear delivery
Architecture, limitations, and operating guidance are documented in language stakeholders can act on.
02 / RESUME
A profile built on verified facts.
A verification-first professional profile. Only the role and engagement channels confirmed in the project brief are published.
Quantitative Developer
Focused on software for quantitative research, market data, risk analytics, and trading workflows.
Fiverr professional profile
Project inquiries are accepted through the verified Fiverr profile linked throughout this site.
Credentials before claims
Employment history, education, certifications, dates, and named tool proficiency are intentionally omitted until confirmed.
03 / SERVICES
Software for decisions that need to hold up.
Five focused ways to turn quantitative requirements into reviewable, maintainable systems.
01
Quantitative Research Systems
- Problem
- Research becomes difficult to trust when notebooks, data, and assumptions drift apart.
- Deliverable
- A structured research environment with repeatable experiments, data contracts, and documented evaluation workflows.
- Client value
- Faster iteration with a clearer path from hypothesis to reviewable evidence.
02
Algorithmic Trading Infrastructure
- Problem
- Strategy logic, execution concerns, and state management are often coupled in fragile scripts.
- Deliverable
- A modular event-driven architecture with explicit interfaces for signals, orders, positions, and controls.
- Client value
- Safer change management and easier testing across research and execution components.
03
Backtesting & Strategy Validation
- Problem
- A polished equity curve can hide leakage, unrealistic fills, omitted costs, and weak benchmarks.
- Deliverable
- A validation workflow covering time-aware data handling, execution assumptions, costs, diagnostics, and scenario tests.
- Client value
- A more defensible understanding of what a simulated result does—and does not—show.
04
Risk Analytics & Portfolio Tooling
- Problem
- Risk signals arrive too late when exposures, limits, and market changes live in separate tools.
- Deliverable
- A portfolio risk layer that consolidates exposures, thresholds, scenarios, and actionable alerts.
- Client value
- More consistent monitoring and clearer escalation when risk conditions change.
05
Market Data & Trading Automation
- Problem
- Inconsistent feeds and manual operations create silent data errors and operational drag.
- Deliverable
- Validated ingestion pipelines and automation with idempotent jobs, observability, and recovery paths.
- Client value
- Cleaner inputs and more reliable operational workflows.
Scope is agreed per engagement. No service includes a promise of profit or financial performance.
04 / SKILLS
Capability areas under review.
These domain-level areas reflect the intended service profile. Specific skills, languages, platforms, and credentials remain unpublished until owner confirmation.
01
Quantitative Engineering
- Research-system design
- Backtesting methodology
- Risk analytics workflows
02
Programming & Data
- Data-pipeline design
- Data quality controls
- Reproducible analysis workflows
03
Trading Systems
- Market-data systems
- Trading automation design
- Portfolio monitoring concepts
04
Infrastructure & Delivery
- Testing strategy
- Technical documentation
- Professional communication
05 / PROJECTS
Technical decisions, made visible.
Three representative case studies prepared for this portfolio demonstrate architecture and validation thinking. They are concepts—not claims of client delivery or production use.
01
Representative Case Study
Event-Driven Backtesting & Research Platform
A research architecture that keeps market events, strategy decisions, execution assumptions, and portfolio accounting separated and testable.
Problem
Vectorized prototypes can hide ordering, fill, and state assumptions that matter once a strategy is evaluated realistically.
Approach
Read case study
Model the workflow as deterministic events with replaceable data, strategy, execution, and accounting components.
02
Representative Case Study
Real-Time Portfolio Risk Monitor
A streaming risk service that turns positions and market updates into explainable exposures, limits, scenarios, and prioritized alerts.
Problem
Risk information loses operational value when positions, prices, limits, and alerts update on different schedules.
Approach
Read case study
Maintain a versioned portfolio state, calculate incremental measures, and separate alert evaluation from notification delivery.
03
Representative Case Study
Volatility Forecasting & Regime Analytics API
A versioned analytics API for volatility estimates and regime probabilities with explicit data lineage, uncertainty, and monitoring.
Problem
A forecast is difficult to use safely when model versions, input windows, uncertainty, and stale-data behavior are hidden.
Approach
Read case study
Serve estimates as versioned analytical products with reproducible training, calibrated outputs, and observable failure modes.
06 / MY BLOG
Notes on research and system design.
Technical essays by Filipe Soares on research engineering, systems architecture, and risk controls.
Published
Preventing Data Leakage in Quantitative Research
A practical framework for preserving the information set, fitting transformations safely, and testing research pipelines for temporal leakage.
August 10, 2026
Published
Designing an Event-Driven Backtesting Engine
How to separate data, strategy, risk, execution, and accounting so a simulator stays deterministic, testable, and honest about fills.
August 10, 2026
Published
Risk Controls for Volatility-Regime Shifts
A layered control design for volatility spikes, thinner liquidity, model uncertainty, and the operational risks of regime transitions.
August 10, 2026
07 / CONTACT
Have a quantitative system to build?
Share the decision, data, constraints, and required deliverable. I’ll help turn the problem into a clear engineering scope.
Start a projectFiverr
(opens in a new tab)
Professional profileLinkedIn
(opens in a new tab)
No public email is displayed because none has been confirmed for publication.