Stockholm · Open to collaboration

Ani

Financial Services Responsible AI Causal Inference

AI/ML Consultant · Financial Services

Building intelligent AI/ML systems for the financial services industry — from causal models and decision engines to responsible, production-grade ML pipelines.

View Work Get in touch ↗

Years in ML

10+

Projects shipped

Selected Work

RL Trading Agent

Python PyTorch RL

Reinforcement Learning

PPO agent trained on historical market data achieving 23% better risk-adjusted returns than buy-and-hold benchmarks. Implements custom reward shaping with drawdown penalties and transaction cost modeling.

View Project ↗

Causal Inference Toolkit

Python Causal ML Statistics

Causal ML

Python library for causal discovery and treatment-effect estimation with a scikit-learn compatible API. Powers A/B test analysis in production across multiple business units.

View Project ↗

LLM Document Intelligence

Python LangChain Azure OpenAI

NLP / LLMs

Multi-document QA system using RAG architecture with custom chunking strategies, achieving 92% accuracy on enterprise document understanding benchmarks across financial and legal domains.

View Project ↗

Ethical AI Dashboard

Python Streamlit XAI

Explainability

Interactive dashboard for auditing ML fairness and explainability using SHAP, LIME, and bias metrics. Used by model governance teams for compliance reporting.

View Project ↗

MLOps Pipeline

Python MLflow Kubernetes

MLOps

Production framework with automated retraining, model versioning, A/B testing, and drift detection — reducing model deployment time from weeks to hours.

View Project ↗

About Me

I'm an AI/ML consultant at Accenture in Stockholm with 8+ years of experience spanning integration engineering, data science, and machine learning. My path from SAP consulting to building RL systems gives me an unusual lens — I think about models the way an engineer thinks about infrastructure: reliability, maintainability, and real-world impact matter as much as benchmark scores.

I did my Master's thesis at Ericsson, combining evolutionary algorithms with reinforcement learning to tackle large state spaces — and later was a TA for AI at KTH. These days I focus on responsible AI, data strategy, and getting ML out of notebooks and into production.

Research

A scalable species-based genetic algorithm for reinforcement learning problems

Knowledge Engineering Review, Vol. 37, e9 · 2022

Extended journal version. Introduces a scalable, species-based evolutionary approach to reinforcement learning, enabling efficient exploration over large state spaces via model encoding and genetic operators. Published in Cambridge University Press. 3 citations.

Distributed Species-Based Genetic Algorithm for Reinforcement Learning Problems

EasyChair · Apr 2021

A novel distributed algorithm combining evolutionary computing with reinforcement learning. Introduces efficient model encoding that enables genetic operators — including crossover in the encoded space — to reduce trainable parameters while achieving comparable performance to DQN and A3C on Atari benchmarks.

A scalable species-based genetic algorithm for reinforcement learning

Preprint · 2021

Early version of the scalable species-based GA work, exploring the combination of evolutionary strategies with deep RL to improve sample efficiency on standard benchmarks.

Machine Learning

Deep Learning

Reinforcement Learning

Evolutionary Algorithms

Data Strategy

Responsible AI

Python

PyTorch

TensorFlow

SQL

Statistics

Data Analysis

Data Science

Visualization

Java

JavaScript

Consulting

Team Leadership

GCP

AWS

Education

Master of Science, Machine Learning

KTH Royal Institute of Technology · 2019 – 2021

B.Engg, Computer Science

BITS Pilani, Dubai · 2012 – 2016

Experience

Jun 2022 – Present

Consultant

Accenture

Leading AI/ML and data strategy initiatives for enterprise clients across the Nordics. Focus areas include responsible AI frameworks, data strategy, and productionising machine learning systems at scale.
Jun 2021 – Jun 2022

Analyst

Accenture

Developed data science solutions and data strategy frameworks for enterprise clients in Stockholm, transitioning from integration engineering into applied ML.
Jan 2021 – Jun 2021

Master Thesis Intern

Ericsson

Investigated the combined use of evolution strategies and genetic algorithms within a reinforcement learning framework to accelerate exploration over large state spaces.
Aug 2020 – Jan 2021

Teaching Assistant

KTH Royal Institute of Technology

Student TA for DD2380 Artificial Intelligence during MSc studies. Ran lab sessions on reinforcement learning, graded assignments, and beta-tested new coding exercises.
Jan 2020 – Dec 2020

Student Intern (SSI)

Scania Group

Participated in Scania's Student Intro programme, exploring applied data and engineering practices in a large-scale automotive manufacturing environment.
Mar 2019 – Aug 2019

Application Development Analyst

Accenture

Designed and maintained integration architectures using SAP PO/CPI. Delivered technical design documentation and Java-based enterprise integrations for clients in Mumbai.
Oct 2016 – Mar 2019

Associate Consultant

Capgemini

SAP PI/XI integration development, enterprise portal builds, business intelligence reporting, and quality assurance across financial services and manufacturing clients.
Feb 2016 – Jul 2016

Full Stack Web Developer

ICBA

Built full-stack web applications using Drupal CMS for the International Centre for Biosaline Agriculture, supporting agricultural research data management in the UAE.