Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

I am a Software Engineer with 4+ years of experience building high-performance ML and information-retrieval systems using Python and C++. I bring a strong grounding in data structures, distributed systems, and large-scale model serving. I have hands-on experience with vector search, token-based retrieval, and GenAI infrastructure. I enjoy collaborating with research teams to translate AI ideas into production systems, performing LLM prompt evaluations, retrieval consistency tests, and model profiling under high QPS scenarios.…I am a Software Engineer with 4+ years of experience building high-performance ML and information-retrieval systems using Python and C++. I bring a strong grounding in data structures, distributed systems, and large-scale model serving. I have hands-on experience with vector search, token-based retrieval, and GenAI infrastructure. I enjoy collaborating with research teams to translate AI ideas into production systems, performing LLM prompt evaluations, retrieval consistency tests, and model profiling under high QPS scenarios.

Shivani M. Latkar

Data Scientist, Full Stack Developer, Web Developer, +2





I am a Software Engineer with 4+ years of experience building high-performance ML and information-retrieval systems using Python and C++. I bring a strong grounding in data structures, distributed systems, and large-scale model serving. I have hands-on experience with vector search, token-based retrieval, and GenAI infrastructure. I enjoy collaborating with research teams to translate AI ideas into production systems, performing LLM prompt evaluations, retrieval consistency tests, and model profiling under high QPS scenarios.…I am a Software Engineer with 4+ years of experience building high-performance ML and information-retrieval systems using Python and C++. I bring a strong grounding in data structures, distributed systems, and large-scale model serving. I have hands-on experience with vector search, token-based retrieval, and GenAI infrastructure. I enjoy collaborating with research teams to translate AI ideas into production systems, performing LLM prompt evaluations, retrieval consistency tests, and model profiling under high QPS scenarios.

Available to hire

I am a Software Engineer with 4+ years of experience building high-performance ML and information-retrieval systems using Python and C++. I bring a strong grounding in data structures, distributed systems, and large-scale model serving.

I have hands-on experience with vector search, token-based retrieval, and GenAI infrastructure. I enjoy collaborating with research teams to translate AI ideas into production systems, performing LLM prompt evaluations, retrieval consistency tests, and model profiling under high QPS scenarios.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Language

Javanese

Advanced

Work Experience

Data Scientist (Contract) at Project 990

May 1, 2024 - Present

Engineered retrieval infrastructure using C++, Python, and AWS SageMaker for LLM-based search across 6M documents; optimized vector indexing and tokenization for 60% latency reduction and 98% semantic accuracy. Built a scalable RAG pipeline (OpenSearch + Lambda + S3) serving real-time embeddings and document retrieval for knowledge graph construction. Prototyped low-latency ML serving modules and batch processing jobs with multi-threaded C++ workers and Python bindings (PyBind11). Collaborated with research teams to translate AI ideas into production systems — LLM prompt evaluation, retrieval consistency tests, and model profiling under high QPS simulation. Designed telemetry dashboards (Power BI / Tableau) to monitor query throughput, memory usage, and precision-recall metrics for serving optimization.

Software Developer at Varroc Engineering Limited

June 1, 2023 - October 15, 2025

Deployed end-to-end React-based applications, leveraging ReactJS, TypeScript, React Context, Playwright, and Jenkins pipelines for CI/CD; supporting real-time automotive analytics and vendor workflows. Developed data models and ETL pipelines integrating PostgreSQL and Snowflake, reducing query latency by 30%. Curated RESTful APIs with authentication, rate-limiting, and encryption for secure transactions. Automated testing and deployment via Jenkins, Docker, and CI/CD workflows, cutting release time by 20%. Enhanced React SPA performance using efficient state management and asynchronous data loading (18% faster render). Engineered Spring Boot microservices for data access and logging, achieving 90% unit test coverage. Applied static analysis and debugging tools to detect vulnerabilities and ensure secure code deployment.