Algorithms Engineer

DESCRIPTION

abra R&D is looking for an Algorithms Engineer!

abra R&D is looking for an Algorithms engineer who will take part in building the a next‑generation agentic analytics platform- a real‑time database optimized for AI agents at scale.

The product focuses on real‑time decision making, causal understanding, and time‑series intelligence over large, high‑dimensional data streams.

This role is about building the algorithmic core of the system- the models and logic that power analytics, forecasting, and decision intelligence in production.

What you’ll do:

Design, develop, and own core algorithms for real‑time analytics
Build causal inference algorithms to model cause‑effect relationships and support decision systems
Develop time‑series models for forecasting, anomaly detection, and pattern discovery on real‑time data
Perform algorithmic research: explore approaches, evaluate tradeoffs, and select optimal solutions
Develop feature engineering and mathematical logic to improve prediction accuracy and stability
Work with high‑dimensional, large‑scale, and event‑driven datasets
Define statistical and mathematical methodologies used across the analytics engine
Integrate algorithms into production systems, with strong focus on performance and reliability
Design experiments and evaluation frameworks to validate algorithm quality
Collaborate closely with engineering teams to ensure models run efficiently in real‑time environments
REQUIREMENTS
Strong background in algorithms, mathematics, and statistics
Hands‑on experience developing production‑grade algorithms
Excellent Python skills for algorithm development

Proven experience with Causal Inference, such as:

Propensity Score Matching (PSM)
Inverse Probability Weighting (IPW)
Difference‑in‑Differences (DiD)
Instrumental Variables (IV)
Regression Discontinuity Design (RDD)
Double Machine Learning (DML) / Causal Forests

Strong experience with Time‑Series Modeling, including:

ARIMA / SARIMA
State Space Models / Kalman Filters
Time‑series anomaly detection and change‑point detection
Deep learning approaches for time‑series (LSTM / Transformers – advantage)

Experience taking algorithms from research to production

Ability to work independently and take full ownership of algorithmic components