AI Program Manager

AI Engineering Leader — Trinetra AILocation: Delhi NCRRole Type: Full-time, LeadershipStage: 0→1 Build and ScaleAbout Trinetra AITrinetra AI is building an AI-native platform for next-generation content creation, intelligence, production and decision-making.The platform brings together Generative AI, multimodal intelligence, video technology, creator workflows, content analytics, production tools and enterprise-grade SaaS/PaaS infrastructure.We are looking for an AI Engineering Leader who can take this vision from 0→1, build the core technology stack, create the engineering team, and scale Trinetra into a robust AI platform.This is not a pure management role. We need a hands-on builder-leader who can move comfortably across AI models, video technology, backend architecture, databases, APIs, cloud infrastructure and frontend applications, and who is willing to prototype or vibe-code when required.Website - https://trinetraai.co/What We Are Looking ForThe ideal candidate combines:DeepTech AI + GenAI + Video Technology + Full-Stack Architecture + SaaS/PaaS + MediaTech + Startup ExecutionWe are particularly interested in people who have already built technology products from an early stage and understand the journey from:Idea → Architecture → Prototype → MVP → Product → Platform → ScaleStartup, founding-team or early-stage engineering experience will be strongly preferred.Key Responsibilities1. Own Trinetra’s AI and Technology ArchitectureDefine and own the end-to-end architecture across:Generative AILLMs and foundation modelsMultimodal AIVision-Language ModelsAI agents and agentic workflowsRAG and knowledge systemsEmbeddings and vector databasesFine-tuning and model adaptationModel orchestrationInference architectureModel evaluation and observabilityAI safety and governanceCost and latency optimisationThe candidate should understand when to build, fine-tune, integrate, orchestrate or use third-party models, rather than simply adding AI APIs to a conventional product.2. Build a Scalable SaaS/PaaS PlatformArchitect Trinetra as a platform, not a collection of disconnected AI tools.Experience should include:Multi-tenant SaaS architecturePaaS architectureAPI-first systemsMicroservicesEvent-driven architectureAuthentication and authorizationRBACDeveloper APIs and SDKsUsage meteringSubscription and billing architectureWorkflow orchestrationEnterprise integrationsObservability and monitoringCloud-native deploymentThe long-term architecture should allow Trinetra capabilities to be consumed through both applications and APIs.3. Deep Understanding of Video TechnologyA critical requirement for this role is strong knowledge of the video technology stack.The candidate should understand:Video ingestion and processingEncoding, transcoding and compressionCodecs and container formatsFFmpeg or equivalent frameworksHLS / DASHAdaptive bitrate streamingCDN architectureVideo storage and asset managementShot and scene detectionFrame-level processingAudio-video synchronizationRendering pipelinesGPU-based processingLarge-scale media infrastructureMetadata extractionVideo workflow orchestrationThe person should understand the technical and infrastructure implications of operating video-heavy AI products at scale.4. Lead Generative Video AI ArchitectureThe candidate should have a deep understanding of the evolving Generative Video AI ecosystem.Relevant areas include:Text-to-videoImage-to-videoVideo-to-videoCharacter consistencyReference conditioningMotion controlCamera controlLip syncVoice generationAI dubbing and localizationVideo inpainting and outpaintingAI editingStoryboard-to-videoScene generationMultimodal content understandingDiffusion and transformer-based architecturesThey should be familiar with leading and emerging model ecosystems such as Veo, Sora, Runway, Kling, Seedance, Hailuo, Luma and comparable open-source and proprietary models.More importantly, the candidate should be able to answer:Which model should be used for which workflow based on quality, speed, consistency, cost and scalability?We want someone capable of building a model orchestration layer so Trinetra can intelligently route tasks across different AI models rather than becoming dependent on a single provider.5. Architect End-to-End AI Video WorkflowsThe candidate should be able to design and scale workflows such as:Script → Scene Breakdown → Storyboard → Character/World Generation → Video Generation → Voice → Music/SFX → Editing → Quality Control → Final OutputThey should understand how to maintain:Character consistencyVisual continuityStyle consistencyNarrative continuityVoice consistencyBrand and IP controlsGeneration qualityVersioningHuman-in-the-loop workflowsInference cost controlProduction reliabilityThe candidate should understand that building an AI studio requires much more than connecting multiple APIs.6. Own Backend and Data ArchitectureThe candidate should be comfortable owning or guiding:Backend servicesAPIsDatabasesData pipelinesModel servicesWorkflow enginesCachingQueuesSearch infrastructureAnalytics infrastructureVector databasesFeature storesData warehousesObject storageStrong knowledge of SQL, NoSQL, distributed systems, vector databases and large-scale data architecture is important.The platform will need to manage large volumes of:VideoAudioImagesScriptsMetadataEmbeddingsModel outputsUser behaviour dataGenerated assets7. Understand Frontend Product EngineeringThe candidate does not need to be a specialist frontend engineer but must understand modern product engineering end-to-end.Relevant experience includes:ReactNext.jsTypeScriptAPI-driven applicationsAI-native user interfacesCopilot and chat interfacesStreaming AI responsesWorkflow applicationsMedia-heavy interfacesReal-time applicationsThey should be capable of making informed architectural decisions across the frontend-backend-AI stack.8. Be Hands-On and Comfortable Vibe-CodingWe want a leader who still builds.The candidate should be comfortable using modern AI-assisted development environments to rapidly create:Proofs of conceptInternal toolsAI agentsAPIsProduct prototypesAutomationWorkflow applicationsTechnical experimentsExperience with tools such as Cursor, Claude Code, Codex, GitHub Copilot or equivalent AI development environments is highly relevant.Vibe-coding should be used as a way to improve experimentation velocity, while maintaining strong engineering standards for production systems.9. Lead the 0→1 JourneyThis is one of the most important requirements.The candidate should have real experience with:Selecting the initial technology stackDesigning architecture from scratchMaking build-vs-buy decisionsBuilding rapid prototypesLaunching MVPsManaging technical debtHiring the initial engineering teamEstablishing engineering practicesIterating with product and usersScaling infrastructure after product tractionManaging cloud and inference economicsWe strongly prefer candidates who have worked in startups, entrepreneurial technology environments or founding teams.10. Build and Lead the Engineering OrganisationThe candidate will help build Trinetra’s engineering team across:AI/ML EngineeringGenerative AI EngineeringVideo AI EngineeringBackend EngineeringFrontend EngineeringData EngineeringMLOpsDevOps / CloudAI Product EngineeringThey should create a culture focused on:Build → Ship → Measure → Learn → ImproveMediaTech Experience — Strongly PreferredCandidates with experience in MediaTech, OTT, streaming, creator technology, gaming, VFX, post-production technology or AI-video startups will be strongly preferred.Relevant experience may include:OTT platformsVideo streamingAI video platformsCreator toolsVideo editingMedia asset managementDigital studiosVFX / virtual productionContent supply chainsLocalization technologyAdTech involving videoContent analyticsThe ideal candidate understands both:How digital media is technically produced and deliveredandHow Generative AI is changing the content production stack.Technical Understanding We ExpectThe candidate should have strong working knowledge across a meaningful combination of:AI / DeepTechLLMsGenerative AIMultimodal AIVision-Language ModelsVideo foundation modelsAI agentsRAGEmbeddingsVector searchFine-tuningModel evaluationPrompt and context engineeringAI inference optimisationBackendPython and/or Node.jsREST / GraphQL / gRPCMicroservicesDistributed systemsEvent-driven architectureAPI architectureQueues and asynchronous processingDataPostgreSQL / MySQLNoSQLRedisVector databasesData warehousesData lakesObject storageData pipelinesCloud & InfrastructureAWS / GCP / AzureDockerKubernetesCI/CDServerless architecturesObservabilityGPU infrastructureAI inference infrastructureFrontendReactNext.jsTypeScriptModern AI-native UX patternsVideo TechnologyFFmpegEncoding and transcodingVideo codecsHLS / DASHCDN architectureGPU video processingMedia pipelinesAsset managementVideo metadataScene and shot processingWhat Will Differentiate a Strong CandidatePreference will be given to candidates who have:Built an AI or DeepTech product from 0→1Built or scaled a SaaS/PaaS platformWorked on Generative AI productsWorked with video foundation modelsBuilt or managed video infrastructureStrong backend and database architecture experienceExperience with multimodal AIExperience with GPU/inference infrastructureExperience orchestrating multiple AI modelsWorked in MediaTech / OTT / creator-tech / AI-videoStartup or founding-team experienceBuilt engineering teamsRemained technically hands-onPersonally shipped production codeStrong product thinkingStrong understanding of AI unit economics