Trianz - Agentic AI Engineer

Company Overview : Trianz is an applied AI solutions company that accelerates customer business transformation through AI powered "Transformation Services as a Software Model". With 25+ years of transforming enterprises, we've evolved to a product-led, platform-driven organization serving global enterprises across Financial Services, Insurance, Healthcare, Hi-Tech, Manufacturing, and other industries.With global presence across 4 continents, our platform portfolio under the unified Concierto brand delivers end-to-end transformations including solutions for Migrate, Manage, Maximize, Modernize, Insights & Agentic AI, and SecOps - delivered through strategic partnerships with leading hyperscales.We're building the premier innovation-led organization in the digital transformation space through AI-first methodologies and data-driven excellence - RevolutionAIzing Transformations.Role Overview : Trianz is taking its market-leading enterprise digital transformation SaaS platform and 20 years of experience in the digital transformation space and creating an AI-native, agentic digital transformation product. This isnt bolting AI onto a legacy SaaS product - this is a complete reimagination. Think Harvey, but for digital transformation.Were looking for a founding-caliber AI engineer to help lead the design and build of this product. Youll help shape the end-to-end architecture and work with cutting-edge AI technologies. This is not a role for someone looking for stability or well-defined processesits for someone who thrives when things are ambiguous and moving fast.Location : Bangalore (Hybrid)Employment Type : Full-timeKey Responsibilities : 1. Agentic AI System Development : - Define and implement the runtime AI architecture for a new AI-native application.- Define and implement the development AI system - fine-tuning pipelines, evaluation frameworks, etc.- Design and build end-to-end agentic AI systems capable of automating enterprise workflows.- Develop systems incorporating orchestration, planning, tool usage, and memory management.- Implement reasoning-driven architectures (agent loops, multi-step execution, decision systems).2. AI/ML Engineering : - Build and optimize knowledge bases and RAG pipelines for enterprise use cases.- Work on fine-tuning, evaluation, and deployment of models.- Contribute to model lifecycle management including monitoring, versioning, and continuous improvement.3. Platform & Architecture : - Define and evolve scalable AI system architectures for enterprise-grade applications.- Work on model deployment strategies, including deploying models without Bedrock, Vertex, and similar managed platforms.4. Product & Enterprise Integration : - Build solutions aligned with enterprise requirements such as security, compliance, and customization.- Collaborate with product teams to translate business problems into AI-driven solutions.- Contribute to building reusable platform capabilities within Concierto.5. Collaboration & Execution : - Work closely with cross-functional teams in a fast-paced, high-ownership environment.- Take end-to-end ownership of critical components from design to deployment.- Drive rapid iteration and experimentation.Ideal Candidate Profile : 1. Experience : - 3 to 7 years of software engineering experience.- 2+ years of hands-on experience building agentic AI systems.- Experience building end-to-end agentic systemscommercial products or mature internal toolsthat have automated and replaced human processes. Were looking for people who have built AI that does real work, not just AI tooling, copilots, or other supplementary features.- Founding/early-stage engineer on an internal agentic venture within a larger company.- Ideally : founding engineer or early-stage senior engineer at an AI or other startup.- Background at vertical AI companies, AI-native startups, or mature Indian startups with serious AI implementation (e.g., Swiggy, Fresh works).2. AI & ML Depth : - Strong understanding of AI system architecture : orchestration layers, tool use, memory, planning.- Experience with retrieval-augmented generation (RAG) pipelines.- Experience with fine-tuned modelstraining, evaluation, and deployment.- Experience with reasoning-focused approaches (chain-of-thought, tree-of-thought, agent loops).- Understanding of the AI system lifecycle : evaluation, monitoring, versioning, continuous improvement of model-driven systems.- Ideally : familiarity with model deployment below the managed-service layer (e.g., deploying on GPUs with inference servers like vLLM or Triton, not just calling Bedrock/Vertex endpoints).- Ideally : pre-LLM AI/ML experience (classical ML, NLP, search, recommendation systems).3. Enterprise & Product : - Experience building enterprise software products with real requirements around security, compliance, private/on-prem deployment, and customer-level customization.4. Mindset & Fit : - Motivated by fast-moving, ambiguous environments with high ownership.- Not someone looking for a comfortable, well-structured big-company role.- Comfortable making architecture and product decisions with incomplete information.Nice to Have : - Domain experience in VMware, cloud migrations, application modernization, ITSM, and/or cloud infrastructure management.- Experience using AI coding agents (e.g., Claude Code, Codex) as part of production engineering workflowsnot just autocomplete, but AI handling architecture, planning, documentation, testing, and code reviews with meaningful autonomy.What we mean by interesting AI experience : This is the part of your background we look at most carefully, so it's worth being upfront about what we're looking for and what we're not. We're not impressed by general software engineering on its own, even strong general software engineering. Python, Kafka, REST APIs, Postgres, Docker, Kubernetes, AWS or GCP these are table stakes for the role, not what makes a candidate stand out. We also don't count what we'd call entry-level AI work : agentic workflows built on Lang Graph or Lang Chain, basic RAG pipelines, prompt engineering, plugging in a vector DB, wrapping the ChatGPT API, or "I built a chatbot." A lot of resumes lead with this kind of work and we understand why it's what most production AI looks like right now but it doesn't tell us much about depth.What does tell us something : - Fine-tuning that goes beyond calling an API. LoRA, QLoRA, RLHF, DPO, full-parameter fine-tuning, instruction tuning. We want to see what you tuned, why, and how it went.- LLM-as-judge evaluation systems where you've thought hard about the hard parts rubric design, calibration against human judgment, handling judge bias, the failure modes you ran into.- Model classes outside the LLM mainstream that you've actually used in the last couple of years : TabPFN, gradient boosting at serious scale, custom transformer variants, diffusion models, structured prediction, mixture-of-experts.- Low-level implementation work : custom CUDA kernels, distributed training, novel architectures, evaluation harnesses you designed from scratch, inference optimization (quantization, speculative decoding, KV cache improvements).- Pretraining or continued pretraining. Any real exposure here is notable.- Reward modeling and preference data construction. (ref:hirist.tech)