Technical Delivery Manager (AI/ML)
ABOUT THE ROLE
The Technical Delivery Manager owns end-to-end program management for Innodata's AI/ML engagements, spanning model development, training and fine-tuning, LLM-based solutions, data pipelines, model evaluation and the wider AI/ML workflow. The role is accountable for keeping what the client expects and what actually gets delivered tightly aligned at every stage, and for surfacing risk early enough that it can still be solved rather than explained.
This is a client-facing position with real exposure. The TDM is the single point of contact for the account, works with stakeholders from day-to-day contacts up to CXO level, and internally connects practice heads, engineering, talent acquisition and operations around one delivery plan. The role is also expected to stay close enough to the technical work to test assumptions, challenge estimates and take a problem apart personally when the program needs it.
KEY RESPONSIBILITIES
Program and Delivery Management
• Own the end-to-end delivery roadmap for AI/ML programs, including model development, training and fine-tuning, LLM implementations and model evaluation workstreams
• Translate client requirements and business goals into structured delivery plans, milestones and success metrics
• Track scope, timelines, budgets and resourcing across multiple concurrent AI/ML projects, flagging slippage early and driving corrective action
• Establish and maintain governance cadences across all active engagements, including status reviews, steering committee updates and sprint or iteration reviews
Client and Stakeholder Management
• Serve as the single point of contact for clients, managing relationships from day-to-day contacts through to CXO-level stakeholders
• Run regular client check-ins, business reviews and escalation calls, presenting delivery status, risks and outcomes
• Capture, document and continuously validate client expectations against what is actually being built, closing gaps before they surface as dissatisfaction
• Build trusted adviser relationships that support account growth and renewal
Cross-Functional Orchestration
• Act as the connective tissue between practice heads, engineering, talent acquisition and operations so that everyone is working off the same delivery plan and priorities
• Coordinate staffing and hiring pipelines with talent acquisition so the right people are available in time for project ramp-up
• Work with practice and technical leads to validate solution approaches, technical feasibility and effort estimates before commitments are made to clients
• Partner with operations on resourcing, utilization, invoicing and billing milestones and contractual compliance
Technical Ownership
• Read and challenge project guidelines, annotation taxonomies, evaluation rubrics and workflow designs rather than routing every technical question to the practice teams
• Interrogate delivery and quality data directly, and build or modify the trackers and working tools the program needs
• Troubleshoot technical delivery issues alongside engineers and data scientists, and test the assumptions behind estimates and solution designs
Risk, Quality and Proactive Issue Management
• Identify delivery risks across technical, resourcing, scope and timeline dimensions early, and drive mitigation before they escalate
• Anticipate client concerns from program signals such as slipping milestones, performance issues or resourcing gaps, and act ahead of formal escalation
• Own issue and escalation management end to end, coordinating the right internal teams and communicating transparently with the client throughout
• Drive continuous improvement in delivery processes, templates and playbooks based on lessons learned across engagements
Reporting and Governance
• Maintain accurate, real-time visibility into program health for internal leadership and client stakeholders
• Prepare and present executive-level dashboards and reviews to both client and internal leadership
• Ensure contractual SLAs, deliverable timelines and commercial commitments are tracked and met
CANDIDATE PROFILE
Education
• Bachelor's degree in engineering, computer science or a closely related technical discipline
• Postgraduate degree preferred. An MBA or equivalent is an advantage but not required where delivery experience compensates
Experience
• 8+ years in technical program or delivery management
• 2 to 3 years minimum specifically managing AI/ML, data science or data engineering programs
• Client SPOC ownership. Demonstrated experience as the primary client-facing point of contact for enterprise or CXO-level stakeholders
• Concurrent program ownership. Proven ability to run multiple cross-functional programs at once in a matrixed environment spanning engineering, practice teams, hiring and operations
• Services delivery model. Experience in IT services, consulting or an AI/ML solutions provider, where delivery is to a client rather than to an internal product team, is preferred
Certifications (Required)
• Project and program: PMP, PgMP or PRINCE2; OR
• Process: Lean Six Sigma Green Belt, consistent with the certification ladder used for IKSI senior operations roles
Preferred
Technical: Cloud or MLOps certification from AWS, Azure or Google Cloud
Technical Skills
• AI/ML lifecycle. Strong working knowledge of model training and fine-tuning, LLM-based solution delivery, model evaluation and MLOps concepts
• Personal technical working capability. Able to read and challenge guidelines and workflow designs, interrogate delivery and quality data directly, build or modify trackers and utilities, and troubleshoot technical issues alongside engineers rather than delegating every technical decision
• GenAI delivery exposure. Direct exposure to LLM and generative AI delivery, including RAG pipelines, fine-tuning, agentic workflows and evaluation frameworks, is preferred
• Delivery tooling. Practical use of Jira, Asana, MS Project or Confluence, plus reporting and dashboarding tools
• Executive communication. Able to translate technical detail into business impact and present credibly to CXO audiences
