// AI Engineer & Full-Stack Developer

Vikramadithya
Baddam

Building intelligent systems at the intersection of ML, GenAI & scalable engineering.

srisaivikramadithyareddy.baddam@sjsu.edu +1 408-646-5150 LinkedIn GitHub San Jose, CA
View Experience Get In Touch
vikram@ai-engineer ~ %
❯whoami.json
{
  "role": "AI Engineer",
  "exp": "4+ years",
  "stack": [
    "PyTorch", "LangChain",
    "LLMs", "RAG"
  ],
  "impact": {
    "gmv_boost": "+18%",
    "f1_gain": "+30%",
    "deploy_speed": "+80%"
  }
}
❯▋
4+Years Experience
+30%F1-Score Improvement
+18%GMV Boost
95%Data Error Reduction
01.

Experience

Egen.ai
Software Engineer (AI)
Aug 2023 – Dec 2025
  • Developed a customer intent prediction pipeline leveraging LSTM and XGBoost, deployed via Kafka and AWS Lambda, increasing targeted promotion effectiveness by 22%.
  • Built a scalable recommendation system using collaborative filtering and deep neural networks with PyTorch and FAISS, boosting GMV by 18% and CTR by 12%.
  • Developed a SpaCy + Transformer-based NLP pipeline to extract indicators from cyber threat intelligence feeds, increasing classification F1-score by 30%.
  • Automated synthetic data generation using GPT-2 and back-translation, enhancing recall by 15% for underrepresented entities.
  • Managed interns for POCs, authored design documentation for smooth collaboration and successful demo results.
SpringML
Data Engineer
Aug 2022 – Jul 2023
  • Deployed an ETL pipeline migrating 8,000 daily pH records to BigQuery with quality checks, reducing data errors by 95% and cutting costs by 20%.
  • Secured Cloud Run outbound traffic via VPC egress with Cloud NAT and static IP for consistent data migration.
  • Designed an optimized BigQuery schema reducing SQL query run time from 2s to 200ms and decreasing data movement by 99.9999% for a 500K row table.
  • Optimized ML pipelines using GCP services (Pub/Sub, Dataflow, Dataproc), enhancing training efficiency and reliability.
  • Implemented GitLab CI/CD pipelines on AWS EKS, slashing deployment times by 80%.
SpringML
Associate Data Engineer
Mar 2021 – Jul 2022
  • Built a voice-enabled CCAI ride scheduling system for NY Paratransit using Dialogflow CX and Java, including ride booking, PIN auth, and recurring trip scheduling.
  • Coupled Genesys Cloud and Dialogflow CX for generative AI features; integrated Google's Knowledge Assist, achieving 90% customer support success rate and reducing live agent assistance by 85%.
  • Built Kafka and Spark Streaming pipelines for chatbot analytics in Looker — intent recognition accuracy 92%, user satisfaction averaging 4.6/5.
02.

Skills

Programming Languages
PythonJavaC/C++JavaScriptTypeScriptGoMySQLPostgreSQLMongoDBRedisDynamoDB
LLMs & GenAI
HuggingFace TransformersOpenAI APILangChainLlamaIndexRAGFAISSChromaAgentic Workflows
ML Frameworks
PyTorchTensorFlowSkLearnNumpyPandasOpenCVNLTKPySparkAirflow
Web & Software
React.jsNode.jsNext.jsVue.jsAngular.jsGraphQLFlaskRESTRedux
Cloud Services
GCPCompute EngineCloud StoragePub/SubFirebaseDataflowTerraformAWS
DevOps & Others
DockerKubernetesGitLab CI/CDKafkaSparkHadoopLinuxBash
03.

Projects

🛋️
Home Furnishing Retailer
Android · Java · Google ARCore · MongoDB · Node.js

Android app for browsing, purchasing, and reviewing furniture with Google ARCore for virtual placement in users' homes. 3D models imported into MongoDB via web crawlers; APIs built with Node.js.

👗
Style My Way
Node.js · MongoDB · React.js · TensorFlow

Multi-portal fashion e-commerce platform with user and vendor portals. Standout feature: a designer portal powered by Neural Style Transfer (TensorFlow) that suggests new designs based on trending fashion.

🧠
NeuralQA — Question Answering
Python · Streamlit · HuggingFace · Dataflow · Vertex AI

Document-level QA system combining dense retrieval (DPR) and extractive QA (BERT). Achieved 92% exact match on a custom dataset.

04.

Education

Indian Institute of Information Technology, Sricity
Bachelor of Technology — Computer Science

Courses: Data Structures, Algorithms, Enterprise Application Development, Machine Learning, Deep Learning, Cyber Security

3.8 / 4.0
Aug 2017 – May 2021