ABBYY’s flagship developer conference in Bengaluru has grown from 70 attendees in 2023 to more than 300 last year. Across these editions, our involvement has grown from offering an intelligent document processing (IDP) hackathon to hosting one of the most anticipated events for the developer community in India. This blog explores how ABBYY’s Developer Conference has evolved and what it says about where Document AI is headed.
The 2026 ABBYY Ascend Developer Conference event is taking place on July 22 and 23 at the Shangri-La in Bengaluru. Visit our event page for more information and to request an invitation.
Three years ago, 70 developers showed up for the first ABBYY Developer Conference in Bengaluru. Last year, we exceeded more than 300 attendees. We're back at it on July 22 and 23 at the Shangri-La, and I've been on the steering committee for all four editions, on stage at every one.
The growth I remember isn’t measured in headcount, though. It’s what happened between sessions.
In 2023, a developer who’d flown in from one of the big global systems integrators (GSIs) stopped me to say an answer I’d posted on the ABBYY community forum had unblocked him months earlier. He’d come with follow-up questions, ready to keep going. Last year, someone showed me a hackathon project he’d built on top of an integration I’d written and listed on the ABBYY Marketplace. He’d taken it somewhere I hadn’t thought to take it myself.
That’s the part that doesn’t show up in an attendee count. You write a forum answer at 11 pm, you ship a connector to the Marketplace, you record an enablement video, and you have no idea who’s on the other end. Then a year later someone walks up and tells you it mattered.
A developer community isn’t the two days in a hotel ballroom. It’s the year of quiet work that makes those two days feel like a reunion.
Our developer conference has changed shape every year, and watching that happen up close has been one of the more satisfying parts of the job.
The first conference had a clear mandate: bring developers together with an intelligent document processing (IDP) hackathon. Participants built solutions with ABBYY Vantage, and the energy in the room made one thing immediately clear: there was a genuine appetite for a dedicated space to dive deep on Document AI. My sessions that year focused largely on evangelizing our products, explaining what purpose-built Document AI could do for teams that had only ever worked with generic optical character recognition (OCR) tools.
The next year, instead of building in isolation, we challenged developers to pitch for value as a group. Teams of one to four competed in a Shark Tank-style format with seven minutes to pitch, two slides, and a judging panel that included ABBYY executives.
This format shift was deliberate. It forced participants to think about value, not just code. A working prototype is impressive, but presenting a clear case on the business value is harder to make. The pitch structure rewarded teams who could connect their technical work to measurable outcomes, and it showcased a wider range of solutions. My own session that year was on retrieval augmented generation (RAG), back when everyone was trying to work out how to ground large language models in real data instead of letting them guess.
The ABBYY Developer Conference in 2025 was the most ambitious yet. Teams from Tech Mahindra, Deloitte, Proservartner, and Telstra built solutions ranging from automated loan applications, invoice processing workflows, and real-life onboarding with generative AI (GenAI) layered in.
What struck me was how rarely the best projects were about ABBYY alone. The interesting work was always at the seams, where our Document AI met someone else’s platform, someone else’s model, someone else’s business problem. That has quietly become the thesis of the whole event.
The questions from the audience reflect a similar maturity. In our first year, our audience wanted to understand what Document AI was. By year three, business leaders and developer engineers in the audience were asking specific questions about combining Document AI with large language models (LLMs) in production environments. The field is visibly maturing in real time, and each observation has quietly informed our next events.
India has become a strategic center of excellence for ABBYY's partner ecosystem and customer base, as well as central to developer talent that drives a significant share of integrations and solutions built on ABBYY products.
Centers of Excellence (CoEs) and Global Capability Centers (GCCs) across India are scaling enterprise automation for organizations worldwide. The ABBYY Ascend Developer Conference exists to meet that talent where it is, and to inspire innovation among that community.
That's also why our events reflect a different type of room that you don't usually get at a developer event.
Our audience includes CoE architects, partner engineers, GCC teams, and customers all in the same sessions, comparing notes and asking each other questions. A partner who’s built something and a customer who needs it often go the extra mile to find each other between sessions. Those hallway conversations are often worth more than the talks.
This year we’re taking the hybrid approach in earnest to inform our agenda. The ABBYY Ascend DevCon format will give participants more flexibility in how to learn and engage, while preserving the in-person aspect that has defined every previous edition.
The honest question every team is wrestling with right now is where ABBYY Document AI should end and where an LLM should begin. The answer matters more than people expect. Lean too hard on a large language model without predefined context and you get a confident wrong answer on an invoice, which in a real workflow is a misposted payment or a compliance miss. This brings far greater risk than simply a spelling error. The interesting engineering is in combining them, so you get accuracy where it's non-negotiable and flexibility where it actually helps.
We will address this challenge directly with sessions on hybrid AI architecture and a roadmap of what is coming from ABBYY. If you work in or around a CoE or a GCC, an ABBYY partner team, or if you’re an ABBYY customer trying to figure out where document automation and GenAI meet, our developer conference sessions are built for you.
Seats are limited.
If you want to be a part of this year’s event, register your interest here: https://digital.abbyy.com/ascend-2026-devcon-register-your-interest/
*McKinsey & Company. Building the foundations for agentic AI at scale, 2026.
