Malas tapi Produktif. Saatnya Pakai AI Agents.

The Dawn of AI Agents: Beyond Chatbots [00:00]

• AI agents are a significant leap from traditional chatbots, possessing the ability to work independently, make decisions, and execute multi-layered tasks.
• They can complete projects from inception to completion without continuous human intervention, offering a potential "game-changer" for companies by boosting productivity dramatically.
• For employees, this technology could signify a shift, with previously secure jobs becoming susceptible to automation.
• The core difference lies in AI agents' ability to connect with and operate other software applications, unlike earlier generative AI like ChatGPT.
• Examples include an AI agent automatically creating presentation slides in Gamma after content is generated in ChatGPT, or sending automated emails or WhatsApp messages.

"AI agent itu selain bisa menjawab pertanyaan kita, mengolah data, tapi dia beneran bisa kayak manusia gitu yang bisa mengerjakan sesuatu yang connect ke software lain."

Understanding AI Agents: Autonomy and Interconnectivity [03:51]

• AI agents can be defined as AI that accesses multiple applications and information sources, working autonomously.
• A practical example is an AI agent acting as a personal assistant, booking flights by checking Google Calendar, browsing airlines or travel platforms, presenting options, and even making the booking and payment.
• These agents can be programmed to analyze initial requests, such as understanding the purpose of a mentoring session, and make decisions on whether it's worthwhile.
• They can also be configured with specific criteria, like requiring certain answers from a potential client before proceeding with a booking.
• Unlike simpler booking tools, AI agents offer interaction and decision-making capabilities.

"Jadi memang perbedaannya adalah benar-benar agen ini seperti ibaratnya ya kayak manusia yang bisa mengerjakan tas gitu, beda sama yang dulu gitu."

Real-World Applications: Media Monitoring and Sales Automation [08:36]

• A compelling use case is media monitoring, where an AI agent can read from various news sources (video, text, audio), summarize content, analyze sentiment (positive, neutral, negative), and generate reports delivered automatically to superiors.
• This capability could render traditional media monitoring roles obsolete.
• Similarly, data analyst roles could be impacted as AI agents can extract, process, and present data autonomously.
• For sales teams, an AI agent can monitor spreadsheets, extract sales data, analyze it, and send daily reports to management at a set time, creating the illusion of extensive human effort.
• This demonstrates the "factory mindset" where the focus shifts from producing output to creating the machinery that produces output.

"Jadi dia akan bisa langsung baca dari sumber-sumber berita yang ada. Betul ya? Jadi kita tentukan kamu baca media apa saja. video, teks, audio bisa tuh."

Complex Use Cases: AI Agents in B2B Sales and Finance [15:07]

• The second level of AI agent capability involves cognitive thinking and decision-making, beyond just tool calling.
• An example involves an AI agent acting as a finance and sales assistant for a B2B company selling fluctuating-price goods like meat or produce.
• The AI agent can handle requests from sales staff regarding discounts for specific clients based on loyalty and other criteria.
• It accesses databases (even spreadsheets), determines pricing, ensures profit margins are maintained, and personalizes pricing for customer loyalty.
• The agent can then automate tasks like sending invoices to finance and sales orders to delivery teams, effectively replacing the work of multiple individuals.

"AI agent lah yang berinteraksi sama orang sales-nya gitu. Dia yang akan ngecekin ke data gitu, database-nya yang bentuknya pun juga masih spreadsheet pun bisa gitu."

Mitigating Risks and Human Oversight [18:46]

• Risks associated with AI agents include incorrect data processing leading to wrong discounts, incorrect suggestions, or sending invoices to the wrong recipients with incorrect amounts.
• To mitigate these risks, AI agents are equipped with "guardrails" to prevent erratic responses.
• A crucial strategy is implementing a "human in the loop" approach, where humans remain the ultimate approvers for critical decisions and validations.
• Many AI agents developed are for backend operations, not direct customer-facing roles, to reduce risk and maintain the human touch in client relationships.
• AI agents can significantly reduce administrative burdens (e.g., 70% of tasks), freeing up sales personnel to focus on client acquisition and relationship building.

"Makanya kenapa si ini enggak kita kasih aja langsung ke customer kan bisa aja kan kayak customer ngomong aja sama si agen ini."

The Genesis of Sinkro: Bridging the AI Agent Gap [22:34]

• Binar Academy's creation of Sinkro, in collaboration with Chinese AI companies, was driven by a desire to understand and adapt the advanced AI agent technology.
• The founders visited China, experiencing the rapid pace of AI development and the culture of interconnectedness among companies.
• They partnered with a Chinese company specializing in AI agents for the financial industry, which had been developing this technology for two to three years prior to its mainstream emergence.
• Sinkro utilizes a platform developed by this partner, allowing for rapid AI agent creation (as little as 15 minutes).
• Binar Academy acquired the platform and is localizing it for the Indonesian market, focusing on integration with tools like spreadsheets, which are prevalent in Indonesia.

"Jadi, jadi satu perusahaan kan jadi mereka apa namanya kita sekarang lagi memodifikasi terus nih platform ini untuk bisa fit ke market Indonesia yang banyak menggunakan spreadsheet gitu."

Embracing Change: The Future of Work with AI Agents [35:36]

• The advice for individuals feeling threatened by AI agents is to embrace the change and learn the technology.
• The analogy used is that of swimming with the waves rather than against them; continuous learning is essential for adaptation.
• The pace of technological advancement requires individuals to "learn how to learn" to stay relevant.
• In fields like programming, AI is already handling a significant portion (90% in the US) of coding, highlighting the need to evolve.
• The ultimate goal is to shift human focus from routine tasks to uniquely human capabilities like creativity and critical thinking.

"Mau enggak mau ya kita harus juga makin cepat untuk beradaptasi kan gitu. Jadi ya skill yang harus dipunya semua orang sekarang adalah belajar cara belajar gitu."

Adoption Timeline and the Power of Sinkro [38:38]

• The widespread adoption of AI agents is predicted for 2026 in the US and 2027-2028 in Indonesia, a slight lag due to development complexities.
• Sinkro aims to simplify AI agent creation, making it accessible to clients who can either train their own teams to use the platform or have Sinkro build agents for them.
• The ultimate goal is for companies to develop their own AI agents, fostering a future where AI agents are integrated into organizational structures, potentially outnumbering human employees.
• Binar Academy's experience shows that AI agents increase productivity, not replace jobs. Their revenue has tripled, and tasks like CV generation have been reduced from 1.5 hours to 5 minutes.

"Ultimately sebenarnya harapannya mereka akhirnya bisa buat sendiri ya. Sebenarnya end goalku ke situ karena end goalku yang aku lihat adalah jadi nanti ya di dalam satu perusahaan organization structure kita, Dok, itu udah manusia AI agent-nya mungkin satu manusia ada tiga gitu ya."