Christiano C.

PROFESSIONAL EXPERIENCE

AI Engineer

Relay Human Cloud | India | Apr 2025 – Present

• Designed and implemented AI-based solutions to streamline HR, compliance, and finance workflows, reducing manual effort and improving operational efficiency.
• Built LLM-powered chatbots and document intelligence systems to enable faster information retrieval and intelligent process automation.
• Trained and optimized AI models on domain-specific documents in collaboration with product and engineering teams to improve response accuracy and relevance.
• Standardized and cleaned data from multiple external service providers to improve data quality, reliability, and accessibility across systems.
• Developed scalable data ingestion and integration pipelines using Mage, consolidating raw data into a centralized master database in Supabase.
• Integrated Apache Kafka for streaming and asynchronous ingestion to support event-driven processing and decoupled system architecture.
• Implemented data validation, schema alignment, and deduplication logic to ensure consistency and integrity across datasets.
• Automated IBV report generation workflows, significantly reducing turnaround time while improving accuracy and repeatability.
• Built operational and analytics dashboards in Metabase to monitor pipeline health, data quality metrics, and business KPIs.
• Monitored and maintained data pipelines, proactively resolving failures and edge cases to ensure high availability and system stability.

Jr. Data Scientist

Valens Datalabs Pvt Ltd | India | Sep 2023 - Mar 2025

• Developed an LLM-powered risk profiling web application that automated and streamlined the manual profiling of entities involved in banking transactions, reducing processing time from 60 minutes to 15 seconds.
• Leveraged Bing Advanced Search API, Mistral 7B-Instruct-v0.2, and Azure services (Virtual Machine, App Service) to enhance the web application’s functionality and efficiency.
• Designed a RAG-based Q&A and learning system for new trainees in a food manufacturing plant, eliminating the need for constant one-on-one human training and significantly reducing training overhead.
• Implemented context-aware responses with source references and timestamped video guides, optimizing the learning process for new employees.
• Utilized GPT-4-32k, Text embedding-3-large, Azure AI Search, and Azure AI Video Indexer to build and deploy the learning system.
• Developed a cultural-accurate English-to-Indonesian subtitle translation API, automating a multi-hour subtitle localization process and reducing translation time to seconds.
• Engineered a two-stage Lambda workflow using AWS API Gateway and Lambda to increase scalability and consistency in media localization, improving the overall quality and speed.
• Used AWS Bedrock (Claude Sonnet 3.5 v2), API Gateway, Lambda, and S3 to deploy the translation service.
• Created a high-performance API for generating transcripts, summaries, sentiment analysis, diarization, PII masking, and politeness detection for customer care call audio files, processing 9000 hours of audio per day.
• Applied Llama 3.1 8b, AWS g4dn.2xlarge instance, xlm-roberta-large-tydip, and pyannote/speakerdiarization-3.1 to build the processing pipeline.