Kailash Dofe profile photo 1 on Tal

Kailash Dofe

AI Engineer (Agentic Systems) @Atlassian | LLM Systems · RAG · Agent Orchestration · GenAI | Open to Forward Deployed Engineer roles

Bengaluru

6+ years of work experience

Atlassian
Oct 20251 yrPresent

Bengaluru, Karnataka, India

As the sole technical decision-maker for a first-of-its-kind AI agent initiative spanning multiple engineering and platform teams, I scoped and shipped a production solution with no existing playbook to follow, partnering directly with support and operations stakeholders, troubleshooting across APIs, auth, and backend systems, and delivering two production AI agent workflows that lifted resolution rates by roughly 10%. To make that sustainable beyond a single project, I designed a modular, skills-based agent orchestration framework, a reusable catalog of skill modules, feature-flagged versioning for safe rollout, and state-machine-driven multi-step diagnostics, replacing brittle, hardcoded prompt chains and cutting both prompt duplication (~30%) and new-workflow shipping time (~40%). On the model side, I ran head-to-head production evaluations across multiple LLM providers and reasoning-effort configurations, benchmarking recall, precision, latency, and response verbosity with an observability platform and a multi-turn evaluation harness over golden datasets, then modeled cost and latency budgets, using parallel and consolidated API calls to hit sub-5-second response times within a lean monthly cost target. I also built a multi-stage LLM pipeline that processed tens of thousands of real, unlabeled support tickets end-to-end (sub-bucketing, intent identification, categorization, and opportunity scoring) with no existing taxonomy to start from, directly informing roadmap and resourcing decisions. Beyond delivery, I authored an internal reference guide on agentic design patterns that was adopted org-wide and mentored several engineers on agentic workflow design.
Oracle
Oct 20212 yrs 3 mosDec 2023

Bengaluru, Karnataka, India

Took ownership of end-to-end service development, from gathering requirements to designing solutions, developing APIs, implementing security measures, and enhancing user experience, resulting in the successful deployment of 5+ major features. These included impactful functionalities such as importing pre-labeled datasets, benefiting 100+ datasets daily, and supporting CSV input datasets, used by 10+ teams. Played a pivotal role in developing the service architecture document and shaping API specifications, streamlining 20+ APIs for seamless integration while ensuring compliance with OCI guidelines. Contributed to maintaining 99.99% service availability, reducing downtime by 40% through robust design, implementation, and operational excellence. Integrated service health and API observability metrics into monitoring systems, boosting troubleshooting speed by 25% and operational efficiency. Designed and automated alarms for service-critical KPIs, enabling proactive issue detection and reducing response times by 30%. Created comprehensive runbooks and conducted 10+ intra-team sessions, achieving a 50% improvement in on-call monitoring efficiency. Additionally, automated legacy workflows, saving 10+ hours weekly on issue resolution and enhancing overall team productivity. Mentored new joiners to full productivity within 3 months, strengthening team capacity and fostering growth. Developed proof-of-concepts (POCs) on Kubernetes, demonstrating the potential to reduce service startup time by 50%, highlighting opportunities for improved agility and performance.
PayPal
Jul 20201 yr 4 mosOct 2021

Bengaluru, Karnataka, India

Successfully migrated 100+ rules from the ODM rule engine to a more efficient system, reducing processing time by 20% and improving overall performance. Conducted comprehensive load and performance testing with a 99.63% success rate, ensuring system reliability and production readiness. Implemented the Java RuleBook framework using lambda expressions, simplifying rule-based logic and enhancing maintainability by 30%. Designed and deployed a decisive planning framework that improved decision-making accuracy and scalability for 5+ critical use cases. Onboarded the application onto a system capable of handling 10% of live traffic for N+1 version testing, ensuring stability and performance prior to deployment. Developed a functional test framework with 85%+ line coverage, significantly increasing test reliability and defect detection efficiency.
JPMorgan Chase & Co.
Jul 20182 yrs 1 moJul 2020

Mumbai Metropolitan Region

Developed a batch job processor capable of handling millions of pending payments, reducing processing time to under an hour, while designing and implementing in-house OAuth-based security protocols to mitigate significant risks and ensure compliance. Leveraged Kafka for seamless inter-module communication, enhancing system performance and reliability. Improved deployment efficiency by 50% through active involvement in deployment activities using an in-house Jenkins wrapper. Monitored system performance using Grafana, processing 10K+ protobuffs weekly to maintain optimal functionality, and achieved 90%+ test coverage with JUnit to ensure robust and reliable code quality.

Kailash’s AI Stack

claude code
claude code
claude
claude
cursor
cursor
gemini
gemini
grok
grok
openai
openai

Education

Motilal Nehru National Institute of Technology

Bachelor’s Degree, Information Technology

Motilal Nehru National Institute of Technology

20144 yrs2018

Join Kailash and 10k+ other professionals on

Tal Boss

Tal is the app where bosses meet engineers & builders directly, without middlemen

Download Tal Boss

Kailash Dofe on Tal

Headline

AI Engineer (Agentic Systems) @Atlassian | LLM Systems · RAG · Agent Orchestration · GenAI | Open to Forward Deployed Engineer roles

About

I’m a software engineer with 8+ years building and operating production systems, now focused on Generative AI: LLM-powered automation, AI agent orchestration, and applied AI solutions. Most of my recent work has been designing and shipping production agent systems, orchestration frameworks, model evaluation pipelines, and automation that measurably improves how teams work. I’m at my best dropped into ambiguous, unfamiliar problems, defining the technical approach from scratch, aligning stakeholders across teams, and moving fast from an open-ended ask to a working, evaluated deployment. Open to connecting on Forward Deployed Engineering, applied AI, or agentic systems.

Location

Bengaluru

AI stack

claude code

claude

cursor

gemini

grok

openai

Work experience

Senior Software Engineer at Atlassian, Oct 2025 - Present · 1 yr, Bengaluru, Karnataka, India, As the sole technical decision-maker for a first-of-its-kind AI agent initiative spanning multiple engineering and platform teams, I scoped and shipped a production solution with no existing playbook to follow, partnering directly with support and operations stakeholders, troubleshooting across APIs, auth, and backend systems, and delivering two production AI agent workflows that lifted resolution rates by roughly 10%. To make that sustainable beyond a single project, I designed a modular, skills-based agent orchestration framework, a reusable catalog of skill modules, feature-flagged versioning for safe rollout, and state-machine-driven multi-step diagnostics, replacing brittle, hardcoded prompt chains and cutting both prompt duplication (~30%) and new-workflow shipping time (~40%). On the model side, I ran head-to-head production evaluations across multiple LLM providers and reasoning-effort configurations, benchmarking recall, precision, latency, and response verbosity with an observability platform and a multi-turn evaluation harness over golden datasets, then modeled cost and latency budgets, using parallel and consolidated API calls to hit sub-5-second response times within a lean monthly cost target. I also built a multi-stage LLM pipeline that processed tens of thousands of real, unlabeled support tickets end-to-end (sub-bucketing, intent identification, categorization, and opportunity scoring) with no existing taxonomy to start from, directly informing roadmap and resourcing decisions. Beyond delivery, I authored an internal reference guide on agentic design patterns that was adopted org-wide and mentored several engineers on agentic workflow design.

Member Of Technical Staff at Oracle, Oct 2021 - Dec 2023 · 2 yrs 3 mos, Bengaluru, Karnataka, India, Took ownership of end-to-end service development, from gathering requirements to designing solutions, developing APIs, implementing security measures, and enhancing user experience, resulting in the successful deployment of 5+ major features. These included impactful functionalities such as importing pre-labeled datasets, benefiting 100+ datasets daily, and supporting CSV input datasets, used by 10+ teams. Played a pivotal role in developing the service architecture document and shaping API specifications, streamlining 20+ APIs for seamless integration while ensuring compliance with OCI guidelines. Contributed to maintaining 99.99% service availability, reducing downtime by 40% through robust design, implementation, and operational excellence. Integrated service health and API observability metrics into monitoring systems, boosting troubleshooting speed by 25% and operational efficiency. Designed and automated alarms for service-critical KPIs, enabling proactive issue detection and reducing response times by 30%. Created comprehensive runbooks and conducted 10+ intra-team sessions, achieving a 50% improvement in on-call monitoring efficiency. Additionally, automated legacy workflows, saving 10+ hours weekly on issue resolution and enhancing overall team productivity. Mentored new joiners to full productivity within 3 months, strengthening team capacity and fostering growth. Developed proof-of-concepts (POCs) on Kubernetes, demonstrating the potential to reduce service startup time by 50%, highlighting opportunities for improved agility and performance.

Software Engineer 2 at PayPal, Jul 2020 - Oct 2021 · 1 yr 4 mos, Bengaluru, Karnataka, India, Successfully migrated 100+ rules from the ODM rule engine to a more efficient system, reducing processing time by 20% and improving overall performance. Conducted comprehensive load and performance testing with a 99.63% success rate, ensuring system reliability and production readiness. Implemented the Java RuleBook framework using lambda expressions, simplifying rule-based logic and enhancing maintainability by 30%. Designed and deployed a decisive planning framework that improved decision-making accuracy and scalability for 5+ critical use cases. Onboarded the application onto a system capable of handling 10% of live traffic for N+1 version testing, ensuring stability and performance prior to deployment. Developed a functional test framework with 85%+ line coverage, significantly increasing test reliability and defect detection efficiency.

Software Engineer at JPMorgan Chase & Co., Jul 2018 - Jul 2020 · 2 yrs 1 mo, Mumbai Metropolitan Region, Developed a batch job processor capable of handling millions of pending payments, reducing processing time to under an hour, while designing and implementing in-house OAuth-based security protocols to mitigate significant risks and ensure compliance. Leveraged Kafka for seamless inter-module communication, enhancing system performance and reliability. Improved deployment efficiency by 50% through active involvement in deployment activities using an in-house Jenkins wrapper. Monitored system performance using Grafana, processing 10K+ protobuffs weekly to maintain optimal functionality, and achieved 90%+ test coverage with JUnit to ensure robust and reliable code quality.

Education

Bachelor’s Degree, Information Technology, Motilal Nehru National Institute of Technology