Kumar Saket

Software Engineer

Bengaluru, Karnataka, India10 yrs 8 mos experience

Highly StableAI Enabled

Key Highlights

Over nine years of experience in AI and data engineering.
Led development of AI observability solutions at TruEra.
Architected data pipelines for enterprise-scale applications.

Stackforce AI infers this person is a SaaS-focused Senior Software Engineer with expertise in AI and data engineering.

Contact

Skills

Core Skills

Artificial Intelligence (ai)Data EngineeringMlopsLarge Language Models (llm)Distributed SystemsFull-stack Development

Other Skills

Access Control Lists (ACLs)AgentsAndroid DevelopmentAngularJSApache AirflowApache DruidApache KafkaApache SparkAutomated Software TestingBack-End Web DevelopmentData Build Tool (DBT)Data PipelinesDjangoETCDElastic Stack (ELK)

About

I am a senior software engineer - AI/ML at snowflake. Previously I worked at TruEra, a company that provides AI observability and analytics solutions for enterprise-scale AI applications. I have over nine years of experience in developing and deploying data pipelines, distributed systems, MLOps, and AI tools across various domains and platforms. I am passionate about the latest technology trends and innovations in the field of AI and data science. I enjoy writing code and brainstorming over creative ideas and solutions. I also like to learn new skills and concepts, and I have completed several online courses and certifications on neural networks, deep learning. My goal is to build useful and impactful products that leverage AI and data to solve real-world problems and enhance people's lives.

Experience

Snowflake

Senior Software Engineer - AI/ML

May 2024 – Present · 1 yr 10 mos · Bengaluru, Karnataka, India · Remote

Driving the development of Snowflake's internal enterprise search and AI assistant platform, focusing on delivering cutting-edge AI solutions to enhance employee productivity and knowledge discovery.
Architected and currently developing robust data engineering pipelines (Airflow, Snowflake, dbt) to process and index millions of internal documents with daily updates for enterprise-wide search.
Designed and implemented a high-performance semantic retrieval layer leveraging ColBERT, achieving a significant reduction in query latency from ~6 seconds to under 1 second.
Engineered the core backend infrastructure for the Enterprise Search platform, supporting ~1K daily requests and integrating key features like Role-Based Access Control (RBAC), LLM Gateway, Access Control Lists (ACLs), and comprehensive ML observability.
Leading the development of a multi-agent framework within the Employee Assistant, successfully automating the creation of over 100 daily IT tickets, improving operational efficiency.
Driving performance optimization and enhancing product competitiveness for Snowflake’s Container Runtimes through detailed benchmarking against industry leaders such as SageMaker and Databricks.
Orchestrating the product's readiness for public preview by establishing automated testing protocols, streamlining reporting mechanisms, and creating comprehensive quickstart guides and documentation.

Artificial Intelligence (AI)AgentsRetrieval-Augmented Generation (RAG)Data Build Tool (DBT)Snowflake CloudData Engineering

Truera

Senior Software Engineer

Sep 2022 – May 2024 · 1 yr 8 mos · Bengaluru, Karnataka, India · Hybrid

(Acquired by: Snowflake - https://www.snowflake.com/blog/snowflake-acquires-TruEra-to-bring-llm-ml-observability-to-data-cloud/)
At TruEra, I led and contributed to projects centered on AI observability and the enhancement of AI applications at an enterprise scale. My key contributions include:
MVP Development: Built and launched an MVP for AI observability, designed to monitor and manage production data effectively across vast scales. This tool has significantly improved the reliability and performance insights of AI systems in production environments.
https://truera.com/resources/truera-launches-truera-ai-observability-first-full-lifecycle-ai-observability-solution-covering-both-generative-and-traditional-ai/
Rule-Based Alert System: Led the development of a rule-based alert system capable of integrating with multiple communication channels including Email, Webhook, and Slack. This system supports robust real-time notifications, enhancing operational responsiveness.
LLM Observability Platform: Engineered a comprehensive observability platform for large language models (LLMs) that leverages open-source tools to facilitate enterprise-scale deployments of generative and predictive AI. This platform is crucial for improving the transparency and effectiveness of AI models in real-world applications.
https://www.globenewswire.com/en/news-release/2024/03/13/2845472/0/en/TruEra-Launches-LLM-Evaluation-and-Observability-that-Combines-Leading-Open-Source-Software-and-Massive-Enterprise-Scale.html

JavaApache DruidGrafanaPythonMLOpsLarge Language Models (LLM)+6

Nference

Senior Software Engineer

Feb 2019 – Aug 2022 · 3 yrs 6 mos · Bengaluru Area, India

Among the first 10 engineers at company.
Building softwares to enable early disease detection:
https://www.cipherbio.com/data-viz/organization/Anumana
https://www.anumana.ai/
Built a generic distributed system to allow Asynchronous jobs to execute, along with caching the results. It also included features like, rate limiting the requests, book-keeping and notification.
Lead development of a central distributed Notifications system. This supports communications across multiple channels. Currently being used by multiple apps at nference.
Built an ETL pipeline to extract patients data from (Electronic Health Records)EHR. Process multi TBs of data daily and make it accessible in a common format - to be consumed by other services like Spark, ElasticSearch etc..
Written Event-driven, reliable and high performant backend system to support core business(collections app) at nference which serves more than 100K API requests daily.
A lot of features were ported(coded in GoLang) from a stable code base(earlier written in Python) and response times and performance were optimised by 20X(for few of the cases).
Worked on creating an event-driven, reliable and high performant backend system to show enriched results for a given query. This involved using hypergeometric and Gene Set Enrichment Analysis(GSEA) - which were CPU intensive tasks.
Built an app to do competitive analysis of authors and companies based on text data spanning across multiple corpus and multi thousand of documents/corpus.

Representational State Transfer (REST)GolangApache AirflowApache KafkaPython (Programming Language)Distributed Systems+8

Boxx.ai | ai for e-commerce

Senior Software Engineer

Jul 2017 – Feb 2019 · 1 yr 7 mos · Bengaluru Area, India

(Acquired by: Netcore - https://boxx.ai/)
Worked on Google AdWords, to create automated custom campaigns based on Boxx.ai AI algorithms.
Worked on integration of plugins like Magento(1.9 and 2.0) as well as Shopify to the client side application, http://app.boxx.ai
Collaborate with delivery as well as development team to get plugins like Magneto, Shopify to the market. Worked on building few APIs, mostly to automate data sync and installation flow of plugins on the client store.
Implemented billing module for auto invoicing and receipt generation, this included integration with Razorpay for payment clearance. Alert for withstanding invoices was also implemented.
Worked on building Dashboards which will show detailed metric for performance of AI based recommendation to clients.
Super Reporting - Processing million of rows of records per minute and showing analytics based on the results.
Improving UI/UX of http://app.boxx.ai
Tech used: Django, React, MongoDB, MySQL, Elasticsearch