Codi did get a Shark Tank deal. Robert Herjavec offered Pillar Learning $500,000 for 25% of the company, and the founders accepted. But the most recent evidence from Pillar Learning tells a more complicated story than the usual “Shark Tank success” update.
The company is still operating a website for Codi, still maintains a product page and user manual, and still describes Codi as its flagship storytelling robot. The catch is that Codi is currently listed as sold out.
That distinction matters because a number of Shark Tank update sites now describe Codi as either dead, inactive, or definitively out of business. The company's own website does not support that conclusion.
Codi entered Shark Tank asking for $500,000 for 10%
Pillar Learning appeared on Season 12, Episode 8 of Shark Tank, which aired December 11, 2020. The company was represented by founders Dayu Yang, Chris Oslebo and William Mock, who pitched Codi as a screen-free educational robot for young children. ABC's own episode listing identifies Codi as a Pillar Learning product and describes the company as developing AI-enabled educational storyteller toys. (abc.com)
The original pitch was straightforward: Codi was designed to occupy the space between a traditional stuffed toy and a smart speaker. It could tell stories, play songs, guide children through routines and deliver educational content, while parents controlled the experience through an accompanying app.
The founders asked for $500,000 in exchange for 10% equity, implying a $5 million valuation.
At the time of the pitch, the business had generated about $260,000 in sales, according to contemporary Shark Tank coverage. The product was priced around $149.99 at retail, while the company's stated manufacturing cost was roughly $25.76 per unit. (sharktankblog.com)
Those numbers were part of what made the pitch difficult. Codi was a physical consumer product, but the company was also positioning it as an education and technology platform. That meant Pillar Learning needed more than a good-looking toy. It needed distribution, repeat engagement and enough scale to make the economics work.
Robert Herjavec came back with a $500,000 offer
The pitch appeared to be heading toward a no-deal before Robert Herjavec returned with an offer.
The deal reported by Pillar Learning was $500,000 for 25% of the company. That was substantially more equity than the founders had originally offered, cutting the implied valuation from their $5 million ask to $2 million. Pillar Learning publicly announced the deal in December 2020, saying Herjavec had offered the $500,000 investment to help expand Codi's reach. (prnewswire.com)
That is the part of the Codi story that is often flattened by later update articles. The televised agreement and the final investment are not necessarily the same thing on Shark Tank. Deals announced on camera are subject to due diligence and can change or fail after filming.
In Codi's case, there is no strong primary-source evidence available today establishing that Herjavec's $500,000 investment actually closed.
Pillar Learning's own post-show announcement said it had “landed” the deal, and its current website still says that Codi's founders “reeled in a deal” on Shark Tank. (prnewswire.com) But that is different from documentation showing that the transaction was completed and funded.
That distinction is worth preserving rather than filling the gap with speculation.
Codi initially positioned itself as more than a toy
The product itself was unusually ambitious for a children's robot.
Codi could play a library of stories, songs and lessons, while the parent app allowed adults to control volume and content, review what a child had listened to and establish guided routines. The company also built in voice messaging between parents and children.
Pillar Learning's current user manual still describes Codi as a connected smart toy with more than 200 songs, stories and lessons, along with a companion parent app. (pillarlearning.com)
The current product page similarly describes Codi as an AI-enabled toy with more than 200 curated songs and stories, personalized content and screen-free interaction. (pillarlearning.com)
That is significant because it shows the product was not simply abandoned immediately after the television appearance. Pillar Learning continued maintaining Codi's software, documentation and product infrastructure.
The company also says Codi sold out during its first week after appearing on Shark Tank and that tens of thousands have been sold in the United States. Those are Pillar Learning's own claims, rather than independently audited sales figures. (pillarlearning.com)
Pillar Learning is still presenting Codi as a live product
This is where the current evidence becomes more interesting.
Pillar Learning's website is still live and still built around Codi. Its product page currently lists the robot at a discounted price of $124.99, reduced from $149.99, but marks the product “Sold out.” (pillarlearning.com)
The international product page also lists Codi as sold out. It describes a robot with more than 200 curated stories, songs and lessons, offline playback and parental controls. (pillarlearning.com)
The company also maintains a dedicated Codi content page, where it says the library includes more than 200 songs, stories and lessons and describes the content as continuing to expand. (pillarlearning.com)
There is even a current Codi user manual hosted on Pillar Learning's website. It describes the device, its cloud connection, parent application and software functionality. (pillarlearning.com)
Taken together, those are much stronger signals than a directory labeling the business “out of business.”
They do not, however, establish that Codi is actively shipping new inventory today.
So is Codi out of business?
The safest answer in August 2026 is: Codi's current commercial status is unclear, but Pillar Learning has not publicly presented itself as closed.
Some third-party Shark Tank databases now classify Pillar Learning's Codi as out of business or “not operating / unclear.” (sharktankcompanies.com) A March 2026 BGR report likewise said the Codi robot was sold out and that there was no recent company announcement explaining when it would return. (bgr.com)
But those reports conflict with what Pillar Learning itself continues to publish.
The company's site remains operational. Its Codi product pages remain accessible. The product is still described in present tense, and the company still provides product documentation and a way to contact it about Codi. (pillarlearning.com)
That leaves a narrower, more defensible conclusion: Codi is not demonstrably dead, but it is also not demonstrably operating as a normally stocked retail product in 2026.
The “sold out” label could mean inventory has temporarily run out. It could also reflect a business that has largely stopped manufacturing while keeping its online infrastructure alive. The available primary-source evidence does not tell us which.
The Shark Tank deal is also less certain than the episode makes it look
There is another unresolved piece.
Pillar Learning publicly announced Herjavec's $500,000 deal after the episode, and multiple Shark Tank references continue to report the agreement as $500,000 for 25%. (prnewswire.com)
But there is no readily available primary-source confirmation that the investment actually closed after due diligence.
That means it would be misleading to write that Herjavec definitely invested $500,000 in Pillar Learning simply because the television segment ended with a handshake.
It would be equally misleading to state that the deal definitely fell apart.
The evidence supports the more limited statement: Herjavec made the $500,000-for-25% offer, the founders accepted it on the show, and Pillar Learning publicly announced the deal afterward. Whether the investment ultimately closed is not clearly documented in the sources available today.
That is probably why Codi produces such contradictory “Shark Tank update” answers online.
Codi's story ended up being quieter than its Shark Tank moment
Codi had several things going for it in 2020. It had a clear consumer problem, an attractive physical product and a pitch that fit the growing interest in reducing children's screen time.
Pillar Learning also built something more substantial than a one-off gadget. Codi had cloud connectivity, a parent app, curated educational content and software that could theoretically evolve after the hardware was sold. The company continues to describe that ecosystem today. (pillarlearning.com)
What is missing is evidence of the kind of post-Shark Tank expansion that would make Codi an obvious breakout success.
There is no current indication of broad retail availability. The company's own store says sold out. There is no reliable current revenue figure establishing how large Pillar Learning became, and the often-repeated “$2.5 million net worth” figures found on Shark Tank aggregator sites are estimates rather than company-reported financials.
So the cleanest Codi update is not that the company became a major Shark Tank success, nor that it definitively disappeared.
Pillar Learning is still online, Codi still exists as a supported product, and the company still promotes the robot. But as of August 2026, Codi is sold out, its future availability is unclear, and the final status of Robert Herjavec's televised $500,000 investment is not independently established.
For a product that once arrived in the Tank asking to become the next big children's technology platform, that uncertainty is the real update.
