Pratik Gehlot

Software Engineer

San Diego, California, United States1 yr 2 mos experience

AI EnabledAI ML Practitioner

Key Highlights

Expert in AI and Machine Learning solutions.
Proven track record in backend systems and real-time data processing.
Strong experience in developing scalable applications.

Stackforce AI infers this person is a Backend-focused Software Engineer with strong expertise in AI and Data Engineering.

Contact

pgehlot@ucsd.edu LinkedIn

Skills

Core Skills

Backend SystemsMachine LearningData Engineering

Other Skills

AWSAirflowAmazon RekognitionAmazon TextractAmazon Web Services (AWS)AnalyticsApache KafkaArtificial Intelligence (AI)C++Computer VisionComputer visionConvolutional Neural Networks (CNN)Data AnalysisData ScienceDeep Learning

About

Software Engineer | AI & Machine Learning | Computer Vision | Backend Systems - Innovative problem-solver passionate about building scalable AI solutions and optimizing system efficiency - Experienced in machine learning, deep learning, and computer vision, with a strong focus on real-time inference, model optimization, and deployment - Skilled in full-stack development & backend systems, integrating AI with scalable architectures, cloud computing, and automation - Driven to leverage AI for real-world impact, from autonomous robotics and debugging automation to document processing and e-commerce platforms - Collaborative team player, thriving in fast-paced environments and delivering results - Open to connecting! Excited to contribute to AI-driven projects and explore opportunities in the tech industry 📩 Email: pgehlot@ucsd.edu

Experience

Vera

Software Engineer

Apr 2024 – Jun 2025 · 1 yr 2 mos · San Diego, California, United States · On-site

Implemented chat and story modules for a social media app using Java (Spring Boot) and DynamoDB, enabling realtime communication for 2K+ users.
Designed low-latency backend pipelines with Spring Boot + DynamoDB, leveraging asynchronous processing, connection pooling, and indexed data models to support <100ms response times and seamless horizontal scaling.
Implemented a real-time data handler in C++ to process incoming events such as likes, comments, and viewer counts during live sessions with minimal latency.
Worked on memory and performance profiling of C++ components, reducing app size by 15% and startup latency by 25% across a wide range of Android devices.
Implemented a ML pipeline with Amazon Rekognition to flag inappropriate content, using Lambda, SQS, and CloudWatch for scalable detection, logging, and alerting within milliseconds.

JavaSpring BootDynamoDBC++Amazon RekognitionBackend Systems+1

Scicrunch inc.

Software Engineering Intern/ Graduate Curator

Oct 2022 – Feb 2024 · 1 yr 4 mos · San Diego, California, United States · On-site

Implemented distributed data pipelines using Airflow, Kafka, and on-prem caching to reduce ETL runtimes from 3 days to under 3 hours across 10+ data sources.
Deployed data pipelines on AWS, enabling 5+ teams to ingest and utilize data with 60% faster turnaround, accelerating experimentation and decision-making.
Developed wikiGPT, an internal tool using embedding-based semantic search for code/documentation, boosting developer lookup efficiency.

AirflowKafkaAWSData Engineering

Wurth it india

Machine Learning and Data Science Intern

Apr 2021 – Jun 2021 · 2 mos · India · Hybrid

Dsw | data science wizards

Software Engineer Intern

Nov 2020 – Jul 2022 · 1 yr 8 mos · Mumbai, Maharashtra, India · Hybrid

Implemented a deep learning-based time series model to predict item demand for Bonprix, a leading UK fashion retailer, achieving a 15% improvement in forecast accuracy over traditional methods.
Developed an OCR model using Amazon Textract to extract key information from PAN cards, Aadhaar cards, and passports, automating document processing.
Contributed to the initial infrastructure setup for UnifyAI, a platform designed to accelerate AI deployment from ideation to production in just 3–4 weeks.

Deep LearningOCRAmazon TextractMachine Learning