AI Software Engineer - Autonomous Agents & RAG Pipelines
Resume:
Durga Shivanaga Kerthi Venkata (Sai Pratheek)
Boulder, CO Open to Relocation
Phone: 720-***-**** Email: *****.***@*****.***
LinkedIn: linkedin.com/in/pratheekkv/ GitHub: github.com/sprat-kv/
Summary
AI Software Engineer with 4+ year experience focused on building autonomous agent systems, multi-agent orchestration, and retrieval-augmented generation. Skilled in building secure, scalable AI pipelines that integrate APIs, databases, and external tools into autonomous agent workflows. Multi-cloud certified with expertise in Python, LangChain, LangGraph, and modern AI frameworks. My work spans agent orchestration, long-horizon decision making, graph-based knowledge frameworks, and secure integration with APIs and enterprise data sources.
Education
MS, Data Science : University of Colorado, Boulder
B.Tech, Electronics and Communications Engineering : Jawaharlal Nehru Technological University, IN
Technical Skills
ETL/ELT Data Lakehouse & Warehouse Data Governance Database Modelling API design Microservice Python Go SQL R LangChain LangGraph A2A MCP NLP Pytorch scikit-learn OpenAI SDK OOPs & Data Structures FastAPI Flask Qdrant Pinecone CI/CD (Git, GitHub Actions, Jenkins) Apache Spark Databricks Delta Lake DBT Data Factory EMR SageMaker MSK Kafka Kinesis Flink Docker Kubernetes Amazon Glue Athena Redshift Snowflake Form Recognizer Rekognition Tableau PowerBI QuickSight CloudWatch Grafana
Experience
Software Engineer - AI July 2025– Present
CLD-9 INC Boulder, CO
●Spearheading the development of GraphRAG knowledge framework on Neo4j that supports grounded reasoning, source-checking, and traceable recommendations for supplement effectiveness and clinical evidence review.
●Built a multi-agent orchestration system using LangChain and LangGraph with RAG on Qdrant, enabling agents to route intents, plan and complete analyst workflows with secure, policy-controlled database access.
●Deployed a fully serverless Agentic pipeline for multi-page blood-reports, allowing agents to extract structured insights and complete long-horizon reasoning under a minute, improving processing efficiency 60%.
●Implemented observability, evaluation metrics, and human-in-the-loop review layers that monitor agent behavior, detect drift, enforce safety guardrails, and ensure compliant decision-making.
Software Engineer(AI/ML) - Co-op Oct 2024– May 2025
CLD-9 INC Boulder, CO
●Integrated fine-tuned LLMs into multi-step LangChain agent workflows on AWS Bedrock, improving reasoning speed and decision latency by 40 percent while raising engagement by 18 percent.
●Analyzed customer–AI interaction logs using SQL and NLP to identify failure modes and behavior patterns, feeding insights back into agent prompts, routing logic, and product strategy.
●Deployed scalable server RESTful APIs, webhooks, and automated CI/CD pipelines, serving AI agents to the front-end, streamlining cloud–storefront integrations and reducing manual reconciliation 40%.
Graduate Research Associate Oct 2023– Oct 2024
Human Interactions and Robotics Group (HIRO-CU Boulder) Boulder, CO
●Specialized in diffusion-based generative models for motion planning and object manipulation with industrial robotic arms, optimizing advanced automation workflows.
●Designed a spatial positioning algorithm integrating SAM & CREStereo vision models, translating real-world object locations into robotic co-ordinates with 98% accuracy.
●Generated and evaluated multi-stage motion plans on motion_bench_maker using graph-based search and reinforcement-learning-influenced planning strategies, benchmarking performance across complex scenarios.
ML Engineer Sr. Research Associate Aug 2022 – Aug 2023
CloudThat Technologies Pvt Ltd Bangalore, IN
●Directed a 4-member team to build a real-time fraud detection pipeline, cutting decision latency 60% and mitigating revenue loss from pre-matured claims, saving $5–10K annually in insurance claims.
●Deployed a real-time inference pipeline using AWS Lambda, Rekognition, & SageMaker, orchestrated via Apache Airflow, reducing decision-making time by 60% and streamlining fraud detection workflows.
●Engineered EV telemetry pipeline with Kafka, Logstash, and ElasticSearch, achieving millisecond latency and enabling real-time insights on Grafana.
●Authored 15+ multi-cloud architecture playbooks with Terraform/Kubernetes, reducing deployment effort 35% and doubling hand-off speed.
Data Engineer Research Associate Sep 2021 – Jul 2022
CloudThat Technologies Pvt Ltd Bangalore, IN
Deployed scalable cloud applications & serverless solutions for Real-time ML & Analytics workloads, accelerated business report delivery through efficient processing pipelines implemented Spark SQL & Pyspark.
●Architected Amazon S3 Delta Lakehouse with Databricks/dbt, ingesting 2GB every 15 min across 80+ tables, enabling near real-time analytics.
●Built 12+ SQL models and 20+ PySpark jobs in dbt, boosting refresh speed 30% and eliminating QA defects with automated testing, improving reporting efficiency by 30%.
●Designed 10+ Power BI/Tableau dashboards with RBAC, transaction monitoring, employee analytics and accelerating data-driven decision-making for supply chain optimization.
●Earned A+ performance rating (2021–22) and Client Value Award for proactively identifying and addressing process enhancements.
Research Intern Mar 2021 – Aug 2021
CloudThat Technologies Pvt Ltd Bangalore, IN
Gained hands-on experience across AWS, Azure, and GCP along with DevOps tools, containerization, and automation frameworks. Applied cloud fundamentals to provision scalable infrastructure & deploy applications.
●Provisioned and automated scalable multi-cloud infrastructure with shell scripts, Docker, and Kubernetes, improving deployment efficiency.
●Built low/no-code solutions using Microsoft Power Platform to accelerate client workflows and reduce manual effort.
Projects
ClariQ - Compliance Aware Agentic Chatbot GitHub Link
●A compliance-aware AI chatbot using RAG and multi-agent orchestration to enable secure, policy-driven access to de-identified patient data. Enforced HIPAA/GDPR compliance,
●Built a compliance-aware AI chatbot using RAG and multi-agent orchestration for secure, policy-driven access to de-identified patient data, enforcing HIPAA/GDPR rules and running A/B experiments on retrieval and prompt strategies to improve answer relevance and user trust.
Tools used: OpenAI LLMs, LangChain, LangGraph, FastAPI, PostgreSQL, Qdrant, Python, RBAC.
LedgerIt - Realtime Finance Tracker GitHub Link
●A scalable cloud-native Real-Time Finance Tracker enabling expense monitoring, category-based analytics, secure receipt storage, and real-time data processing and resource consumption alerting using AWS services.
Tools used: AWS (Kinesis, Lambda, EC2, RDS, S3, IAM, CloudWatch), React, Flask, PostgreSQL, Python.
Spatials Estimator GitHub Link
●A spatial positioning algorithm that calculates 6D poses for objects integrating SAM & CREStereo vision models, translating real-world object locations into robotic co-ordinates with 90% accuracy.
Tools used: Python, Computer Vision, Inverse Kinematics, ROS2
Certifications
●AWS Certified Data Engineer Associate Credly
●Microsoft Azure Developer Associate Microsoft Certifications
●Databricks Platform Architect Databricks Academy
Contact this candidate