We are inviting proposals for the inaugural edition of the Great International AI Native Summit (GAINS) 2026.
From the makers of the Great International Developer Summit (GIDS), GAINS is a focused, engineering-first summit for teams building and operating AI-native systems in the real world.
📍 Bengaluru, India
📅 09–10 December 2026
This is not a general AI conference.
AI is easier than ever to prototype. It is much harder to make intelligent systems reliable, governable, observable, scalable, and operationally trustworthy in production.
GAINS is about that entire journey: from the first steps of adopting AI in your engineering workflow, to operating intelligent systems at scale in the real world.
Traditional software systems are largely deterministic. AI-native systems are probabilistic. They do not always behave identically under the same conditions. They require new approaches to orchestration, evaluation, observability, supervision, governance, reliability, workflow coordination, and operational control.
These systems are now being built inside real organisations, across fintech, retail, healthcare, logistics, SaaS, telecom, and enterprise platforms, where constraints around latency, cost, compliance, scale, and reliability are non-negotiable.
Whether your team is just beginning to integrate AI into your engineering practice, or already running intelligent systems in production, GAINS is where engineers come to learn from real experience, share hard-won lessons, and figure out what actually works.
🔍 What We Are Looking For
We are interested in sessions grounded in real engineering work across AI, product, and platform systems.
- AI systems deployed in real products and workflows
- AI features built into existing applications across frontend, backend, and full-stack systems
- agent systems, orchestration, and long-running workflows
- backend services, APIs, and integrations powering AI-driven applications
- frontend experiences for AI systems, including human review, approval, and correction flows
- production deployments and operational behaviour
- evaluation, observability, and feedback systems
- failure modes, edge cases, and unexpected outcomes
- architecture decisions under real-world constraints
- scaling challenges across latency, cost, and reliability
- human-agent interaction, supervision, and intervention
- governance, compliance, and operational control mechanisms
- AI infrastructure, inference systems, and runtime behaviour
- integration with existing systems, platforms, and data environments
We are especially interested in talks that discuss trade-offs under real-world constraints: latency, cost, reliability, governance, scale, operational complexity, and uncertainty.
🎯 Your talk should leave the audience with practical engineering insight they can apply immediately.
What Strong GAINS Talks Usually Include
We value talks that clearly explain:
- what the core engineering insight or lesson is
- where that insight comes from: production experience, training practice, research, or hands-on teaching
- what constraints or trade-offs shaped the approach
- what engineers can take away and apply immediately
- what commonly goes wrong and how to avoid it
⚠️ For intermediate to advanced sessions, assume the audience already understands the basics. Skip broad introductions and get into the engineering details.
⚠️ GAINS values depth over breadth. Whether your insight comes from building and operating real systems, or from years of teaching, training, and researching how engineers work with AI; what matters is that your content is grounded, practical, and immediately useful to engineers in the room.
🧭 Focus Areas
🧩 AI-Native Applications
How do AI systems actually get used in real products and workflows?
Systems where AI is not an add-on, but a core part of how the product behaves, makes decisions, or interacts with users. This includes building and shipping AI features inside real applications, across frontend, backend, and full-stack systems.
Examples:
- AI-native products and features
- AI features embedded in existing SaaS and enterprise products
- frontend patterns for human review, approval, and correction of AI outputs
- backend APIs, services, and workflows powering AI-driven applications
- retrieval-augmented systems in production
- multimodal systems in real-world use
- AI-assisted and AI-driven workflows
- integration with legacy and enterprise systems
- human-in-the-loop product experiences
- domain-specific AI systems across fintech, healthcare, retail, logistics, and enterprise platforms
🤖 Agents, Context, and Workflows
How do AI systems act, coordinate, and make decisions over time?
This area focuses on agent behaviour in real environments, not just prototypes.
Examples:
- single-agent and multi-agent systems in production
- context engineering and context lifecycle management
- memory architectures and failure modes
- orchestration and workflow coordination
- tool use, MCP, and external system integration
- planning and reasoning under constraints
- long-running and stateful workflows
📊 Production AI Systems
Does the system behave reliably under real-world conditions?
Evaluation, debugging, and operational behaviour of AI systems after deployment.
Examples:
- evaluation systems and feedback loops
- hallucination detection and mitigation
- reasoning traces and system introspection
- observability for AI systems
- production debugging and incident analysis
- drift, regression, and degradation handling
- fallback and recovery mechanisms
- system behaviour under failure and uncertainty
⚙️ Infrastructure and Scaling
Can the system run efficiently, consistently, and at scale?
Serving, inference, and the operational backbone of AI systems.
Examples:
- inference optimization and model serving
- latency reduction and real-time constraints
- GPU/TPU utilization and scheduling
- scaling bottlenecks in production systems
- distributed AI infrastructure
- caching, batching, and routing strategies
- cost-performance trade-offs and FinOps
- deployment architectures across environments
🔐 Trust, Security, and Governance
Can the system be trusted to operate safely and within defined boundaries?
Security, policy enforcement, and operational control of AI systems.
Examples:
- prompt injection and adversarial behaviour
- data access controls and isolation
- agent authorization and permissions
- auditability, traceability, and logging
- governance systems and policy enforcement
- compliance in regulated environments
- operational safety controls
- risk management for autonomous systems
🌱 AI Foundations for Engineering Teams
A curated track for teams at the beginning of their AI adoption journey.
GAINS 2026 includes a curated Foundations track featuring invited sessions designed for engineering teams in the early stages of integrating AI into their workflows. This track covers practical starting points, foundational engineering practices, and lessons from teams who have recently made this transition. The Foundations track is not open for CFP submissions. Speakers for this track are invited directly by the GAINS programme committee.
🔎 Cross-Cutting Theme
👥 Supervising Autonomous Systems
How do humans retain control over systems that can act independently?
This is a core theme that spans multiple focus areas, especially Production AI Systems, Trust, Security, and Governance, and AI-Native Applications.
Examples:
- human-in-the-loop workflows
- approval and escalation systems
- delegation boundaries and control models
- operator tooling and supervisory interfaces
- fallback, rollback, and recovery strategies
- confidence scoring and decision thresholds
- human-agent collaboration patterns
- intervention strategies in live systems
What Is Unlikely To Be Selected
- Introductory sessions without practical engineering grounding
- Generic “AI will change everything” talks
- Product demonstrations without engineering depth
- Pure framework walkthroughs without operational lessons
- Marketing or promotional presentations
- Broad trend summaries without implementation detail
- Sessions that avoid discussing constraints, trade-offs, or real-world behaviour
- Sessions that lack practical engineering grounding, regardless of the speaker's background
⚙️ Submission Guidelines
🆕 Original content only
Proposals should be based on work you have directly done or materially contributed to.
🧠 Depth over breadth
We value focused engineering insight over broad conceptual coverage.
🚫 No marketing content
Vendor-neutral talks are strongly preferred. Product-focused talks must contain substantial engineering depth and operational learning.
🧪 Engineering detail matters
Architecture diagrams, operational metrics, evaluation approaches, debugging workflows, implementation details, and trade-offs are highly encouraged.
💥 Failure and learning matter
We strongly value talks that discuss constraints, operational surprises, redesigns, workflow changes, and lessons learned.
🎓 Educators, trainers, and authors welcome
GAINS recognises that deep engineering insight is not limited to those with direct production experience. Leading educators, trainers, and authors who can deliver practical, engineer-focused content grounded in real teaching experience and technical depth are strongly encouraged to submit.
🎯 Session Formats
The GAINS 2026 Call for Speakers is accepting the following formats. Using this form, you can submit up to 5 talk proposals.
- Long Talk: 60-minute Presentation
- Short Talk: 30-minute Presentation
For international speakers:
We request that you submit only 60-minute session proposals. You are encouraged to submit multiple sessions. Each proposal is evaluated independently, and submitting more high-quality proposals can improve your chances of selection.
🎤 Speaker Support
Selected International Speakers may receive:
- Travel support of up to US$ 2,000
- 2 nights of accommodation
- Full conference access
Note: International speakers are expected to submit multiple 60-minute session proposals. If your employer or organisation is able to cover or co-fund your travel, please indicate this in your submission so we can prioritise support for independent practitioners.
Selected Inter-state Speakers may receive:
- Travel and accommodation support of up to Rs. 20,000
- Full conference access
Selected Local Speakers will receive:
Additional Notes
- Speaker support is discretionary and not guaranteed upon acceptance.
- Acceptance decisions are based on the strength and relevance of the submission.
- All reimbursements will be processed post-event upon submission of valid bills and invoices.
- Reimbursements will be limited to the approved support caps stated above.
- If your participation is supported by your employer or organisation, please indicate this in your submission. This helps us allocate limited support to independent practitioners.
- Only the primary speaker is eligible for speaker support unless explicitly approved otherwise.
- A maximum of one co-speaker is allowed per session. Approved co-speakers must purchase a valid conference pass.
- The organisers are not responsible for cancellation, rescheduling, or fare differences arising from changes to travel or accommodation bookings.
⏳ Deadline to Submit: July 15, 2026
Whether you are just beginning to integrate AI into your engineering practice or already running intelligent systems in production, we would love to hear from you. Submit below at the earliest and help us shape the inaugural edition of GAINS 2026.
Submit below at the earliest and help us shape the inaugural edition of GAINS 2026.