Utkarsh Sinha

Data Scientist

https://www.opentalent.in/utkarsh-sinha

Data Scientist specializing in Quantitative Finance, RAG architectures, and Time Series Analysis.

India

Key Strengths

Extensive practical experience in various data science domains, including time-series forecasting, fraud detection, NLP, and reinforcement learning, demonstrated through numerous personal projects.
Proficiency in Python and related data science libraries, evident from project technologies.
Experience with MLOps concepts (MLflow, Docker, Makefile) and real-time systems (Redpanda, FastAPI) indicates an understanding of productionizing ML models.
Exposure to advanced ML techniques like Graph Neural Networks (TGN), Transformer policies, Conformal Prediction, and Causal Impact.
Strong alignment with the 'Data Scientist' target role due to the breadth and depth of data science and machine learning projects.

Cultural & Operational Fit

Cultural Fit Analysis

The candidate demonstrates a strong passion for data science and machine learning through a diverse portfolio of personal projects. The projects cover a wide range of applications, from finance to agriculture and healthcare, indicating adaptability and a broad interest in applying data science to various challenges. This aligns well with a culture that values innovation, continuous learning, and problem-solving across different domains. However, the lack of team-based project experience or professional work experience makes it difficult to fully assess cultural fit in a collaborative team environment.

Soft Skills & Operational Fit

Insufficient data to assess soft skills or operational fit. The candidate's project descriptions suggest a strong initiative and ability to work independently on complex problems.

AI is analyzing your overall score…

Identifying your key strengths…

Evaluating your skill match against the job requirements…

Assessing your cultural and operational fit

Projects

pulsepredict

May 18, 2026 – Present

Probabilistic multi-horizon time-series forecasting: PatchTST + TFT + Chronos-T5 + conformal prediction + MinT reconciliation + Bayesian CausalImpact

View Project

graphpulse

May 18, 2026 – Present

Real-time streaming fraud detection on heterogeneous transaction graphs: TGN + LightGBM + Redpanda + River ADWIN + SHAP + GNNExplainer

View Project

asymflow-audio

May 15, 2026 – Present

Rank-asymmetric flow matching for raw audio waveforms (SC09) — porting AsymFlow arXiv:2605.12964 to speech

View Project

documind

May 14, 2026 – Present

Multimodal document RAG — ColPali vision-language retrieval + Qwen2-VL-2B QLoRA generation with citation verification. No OCR needed.

View Project

khetsar

May 13, 2026 – Present

Multi-modal satellite Earth observation platform for Indian agriculture — MAE pretraining, U-TAE crop segmentation, yield forecasting, production MLOps

View Project

prism

May 9, 2026 – Present

PRISM: End-to-end Customer Intelligence Platform — uplift-weighted CLV NBA scorer, churn, MMM, A/B | FastAPI + Next.js 15 + NextAuth + MLflow + Docker

View Project

sahayak

May 7, 2026 – Present

Multi-agent cognitive prosthesis for mild cognitive impairment — LangGraph + LanceDB + InsightFace + Whisper + Federated Learning + Flutter

View Project

rl-market-maker

May 3, 2026 – Present

RL Market Maker on Simulated LOB: PPO + IQN + CVaR vs Avellaneda-Stoikov, GLT, CJ baselines. Hawkes order flow, transformer policy, curriculum training.

View Project

deep-hedging-transformer

May 3, 2026 – Present

Deep Hedging with Transformer policy under Rough Bergomi volatility — MTech quant finance project

View Project

Dynamic-Graph-Portfolio

January 28, 2026 – Present

Dynamic-Graph-Portfolio — GitHub repository

View Project

Key Strengths

Extensive practical experience in various data science domains, including time-series forecasting, fraud detection, NLP, and reinforcement learning, demonstrated through numerous personal projects.
Proficiency in Python and related data science libraries, evident from project technologies.
Experience with MLOps concepts (MLflow, Docker, Makefile) and real-time systems (Redpanda, FastAPI) indicates an understanding of productionizing ML models.
Exposure to advanced ML techniques like Graph Neural Networks (TGN), Transformer policies, Conformal Prediction, and Causal Impact.
Strong alignment with the 'Data Scientist' target role due to the breadth and depth of data science and machine learning projects.

Cultural & Operational Fit

Cultural Fit Analysis

Soft Skills & Operational Fit

Insufficient data to assess soft skills or operational fit. The candidate's project descriptions suggest a strong initiative and ability to work independently on complex problems.

Utkarsh Sinha

Key Strengths

Cultural & Operational Fit

Top Skills

Skills

Projects

Key Strengths

Cultural & Operational Fit